- Manuals
- Brands
- Agilent
- Software
- Feature Extraction 12.2
- User Guide
Agilent Feature Extraction 12.2 Software User Guide PDF
Summary of Content for Agilent Feature Extraction 12.2 Software User Guide PDF
Agilent Feature Extraction 12.2
User Guide For Research Use Only. Not for use in diagnostic procedures.
Agilent Technologies
Notices Agilent Technologies, Inc. 2021
No part of this manual may be reproduced in
any form or by any means (including elec-
tronic storage and retrieval or translation
into a foreign language) without prior agree-
ment and written consent from Agilent
Technologies, Inc. as governed by United
States and international copyright laws.
Manual Part Number G4460-90065
Edition Revision A0, January 2021
Agilent Technologies, Inc.
5301 Stevens Creek Blvd.
Santa Clara, CA 95051
Warranty The material contained in this docu- ment is provided as is, and is sub- ject to being changed, without notice, in future editions. Further, to the max- imum extent permitted by applicable law, Agilent disclaims all warranties, either express or implied, with regard to this manual and any information contained herein, including but not limited to the implied warranties of merchantability and fitness for a par- ticular purpose. Agilent shall not be liable for errors or for incidental or consequential damages in connec- tion with the furnishing, use, or per- formance of this document or of any information contained herein. Should Agilent and the user have a separate written agreement with warranty terms covering the material in this document that conflict with these terms, the warranty terms in the sep- arate agreement shall control.
Technology Licenses The hardware and/or software described in
this document are furnished under a license
and may be used or copied only in accor-
dance with the terms of such license.
Restricted Rights Legend U.S. Government Restricted Rights. Soft-
ware and technical data rights granted to
the federal government include only those
rights customarily provided to end user cus-
tomers. Agilent provides this customary
commercial license in Software and techni-
cal data pursuant to FAR 12.211 (Technical
Data) and 12.212 (Computer Software) and,
for the Department of Defense, DFARS
252.227-7015 (Technical Data - Commercial
Items) and DFARS 227.7202-3 (Rights in
Commercial Computer Software or Com-
puter Software Documentation).
Safety Notices
CAUTION
A CAUTION notice denotes a haz-
ard. It calls attention to an operat-
ing procedure, practice, or the like
that, if not correctly performed or
adhered to, could result in damage
to the product or loss of important
data. Do not proceed beyond a
CAUTION notice until the indicated
conditions are fully understood and
met.
WARNING
A WARNING notice denotes a hazard. It calls attention to an operating procedure, practice, or the like that, if not correctly per- formed or adhered to, could result in personal injury or death. Do not proceed beyond a WARNING notice until the indicated condi- tions are fully understood and met.
Patents Portions of this product may be covered
under US patent 6571005 licensed from the
Regents of the University of California.
Technical Support For US and Canada
Call (800) 227-9770 (option 3,4,2)
Or send an e-mail to:
informatics_support@agilent.com
For all other regions
Agilents world-wide Sales and Support
Center contact details for your location can
be obtained at
www.agilent.com/en/contact-us/page.
2 Feature Extraction 12.2 User Guide
In This Guide
Feature Extraction 12.2 User Guide
This User Guide shows you how to set up and run Feature Extraction automatically for a batch of image files and how to extract image files in real time.
1
This chapter introduces the Feature Extraction analysis process and the tasks you can perform with the program to help you with your microarray experiments.
2
This chapter provides instructions on how to set up and run Feature Extraction on a batch of existing Agilent and/or non- Agilent image files. It also shows you how to set up and run Feature Extraction on Agilent image files as they are saved from the scanner in real time.
3
Automatic extraction can take place because grid templates and grid files can be assigned to an image file. This chapter shows you how to create grid files and templates.
4
Another reason why automatic extraction can take place is because a set of Feature Extraction parameters can be assigned to an image file as a protocol. This chapter gives instructions on how to change the parameter values for each step of the Feature Extraction process.
5
Learn how to work with image displays in this chapter. This chapter also includes instructions on working with the frames and panes of the interface display.
3
Acknowledgments
4
Apache acknowledgment
Part of this software is based on the Xerces XML parser, Copyright (c) 1999- 2000 The Apache Software Foundation. All Rights Reserved (www.apache.org).
JPEG acknowledgment
This software is based in part on the work of the Independent JPEG Group. Copyright (c) 1991- 1998, Thomas G. Lane. All Rights Reserved.
Loess/Netlib acknowledgment
Part of this software is based on a Loess/Lowess algorithm and implementation. The authors of Loess/Lowess are Cleveland, Grosse and Shyu. Copyright (c) 1989, 1992 by AT&T. Permission to use, copy, modify and distribute this software for any purpose without fee is hereby granted, provided that this entire notice is included in all copies of any software that is or includes a copy or modification of this software and in all copies of the supporting documentation for such software.
THIS SOFTWARE IS BEING PROVIDED AS IS, WITHOUT ANY EXPRESS OR IMPLIED WARRANTY. NEITHER THE AUTHORS NOR AT&T MAKE ANY REPRESENTATION OR WARRANTY OF ANY KIND CONCERNING THE MERCHANTABILITY OF THIS SOFTWARE OR ITS FITNESS FOR ANY PARTICULAR PURPOSE.
Stanford University School of Medicine acknowledgment
Non- Agilent microarray image courtesy of Dr. Roger Wagner, Division of Cardiovascular Medicine, Stanford University School of Medicine
This software contains material that is Copyright (c) 1994- 1999 DUNDAS SOFTWARE LTD., All Rights Reserved.
Feature Extraction 12.2 User Guide
Feature Extraction 12.2 User Guide
LibTiff acknowledgement
Part of this software is based upon LibTIFF version 3.8.0.
Copyright (c) 1988- 1997 Sam Leffler Copyright (c) 1991- 1997 Silicon Graphics, Inc.
Permission to use, copy, modify, distribute, and sell this software and its documentation for any purpose is hereby granted without fee, provided that (i) the above copyright notices and this permission notice appear in all copies of the software and related documentation, and (ii) the names of Sam Leffler and Silicon Graphics may not be used in any advertising or publicity relating to the software without the specific, prior written permission of Sam Leffler and Silicon Graphics.
THE SOFTWARE IS PROVIDED AS- IS AND WITHOUT WARRANTY OF ANY KIND, EXPRESS, IMPLIED OR OTHERWISE, INCLUDING WITHOUT LIMITATION, ANY WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
IN NO EVENT SHALL SAM LEFFLER OR SILICON GRAPHICS BE LIABLE FOR ANY SPECIAL, INCIDENTAL, INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE, AND ON ANY THEORY OF LIABILITY, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
5
6
Content
1 Working with Feature Extraction 17
Feature Extraction Process 18
Overview of Feature Extraction Capabilities 21
Feature Extraction program window 21
Feature Extraction algorithms for data analysis 23
Extraction results 24
What You Can Do to Set up and Run Automatic Batch Extraction 25
Setup 25
Run extraction 26
What You Can Do with Images Before Extraction 27
View image and change image display 27
Find grid and spots on the image manually 27
What You Can Do with Protocol Algorithms 29
What you can do with the protocol steps 29
What the protocol steps do for you 29
What You Can Do with.shp Files of Extracted Images 33
What You Can Do with Extraction Results 34
2 Extracting Microarrays Automatically 35
Setting up eArray Login for Automatic Updates 36
Setting Up and Editing Projects 39
To set up or edit a standard project to run existing image files 39
To set up or edit an On-Time project to run image files in real time 41
To create a project 42
To open a project 43
To save a project 44
Changing Properties for Individual Projects 45
Feature Extraction 12.2 User Guide 7
Contents
To display project properties for individual projects 45
To change general project property settings 48
To select the location for saving and/or sending output files 49
To change the FTP settings to send result files automatically to Rosetta
Resolver 52
To set up to automatically assign protocols 52
To select to use a grid file 54
To select an external DyeNorm list file 54
To select to overwrite previous results 54
To select the incoming image folder for On-Time projects 55
To select a time to terminate real-time extractions 55
To turn off the option to automatically extract XDR images 56
Changing Default Project Properties in Preferences 57
To open the Preferences dialog box 57
To display the General default project properties 58
To select to save the Project Run Summary Report 58
To select to overwrite previous results 59
To select a default folder for a new project 59
To display the default project properties for Output and Data Transfer 60
To select the default location for saving and/or sending output files 61
To change the default FTP settings for automatic data transfer 64
To change default run post processing plug-ins 64
To display the default properties for standard projects 65
To change the default protocol and its location 66
To select to use a grid file 67
To change the Start Run Mode setting 67
To select an external DyeNorm list file 67
To select to create XDR extraction sets automatically 68
To display the on-time default properties 68
To select the default image folder 69
To select a default time to terminate real-time Feature Extraction 69
To select to automatically extract an XDR image pair 70
8 Feature Extraction 12.2 User Guide
Contents
Working With Extraction Sets 71
What is an extraction set? 71
How does the program create an extraction set? 72
To add extraction sets (image files) to the standard project 73
To sort extraction sets 75
To remove an extraction set from the project 75
To display/edit the Extraction Set Configuration table 76
Working With Grid Templates/Files 78
To manage grid templates 78
How Dye Normalization Lists are Associated With Array Designs/Grid
Templates 80
Setting Up and Changing Grid Templates and Files 83
To show or hide the Grid Template Browser 83
To add or change a grid template to an extraction set 83
To add or change a grid file to an extraction set 84
To add or remove a grid template 84
To open grid mode from an extraction set 85
To update array designs/grid templates online (via Agilent eArray) 85
To display or change grid template properties 90
To add or change the default protocol in the grid template 91
To add or change the default one-color protocol in the grid template 91
To change the dye normalization gene list in the grid template 92
To create a new internal dye normalization gene list for a grid template 93
To edit an internal dye normalization list 97
To import external dye normalization list to create a new internal list 98
To export an internal dye normalization gene list 98
To create a new critical probes list for a grid template 99
To import a critical probes list 102
To export a critical probes list 103
To edit a critical probes list 103
Working With Protocols 104
Feature Extraction 12.2 User Guide 9
Contents
To move custom protocols from a previous version to version 12.2 104
To show or hide the FE Protocol Browser 104
To add a protocol to an extraction set 105
To change the protocol used in an extraction set 105
To display or change protocol properties 106
To import or export a protocol to or from the database 107
To remove a protocol from the database 107
Working With QC Metric Sets 108
To update the list of displayed QC Metric Sets 108
To move custom QC metric sets from a previous version to version 12.2 109
To show or hide the QC Metric Set Browser 109
To associate a QC metric set with a protocol 109
To import or export a QC Metric Set to or from the database 110
To remove a QC metric set from the database 110
To refresh the QC Metric Set Browser 111
To view metric set properties 111
Running Feature Extraction 113
To start Feature Extraction 113
To stop Feature Extraction 113
To create a new project with remaining extraction sets if run fails 114
Displaying Results 115
To display the text result file in Microsoft Excel 115
To display the Project Run Summary Report 117
To display the QC Report 119
To print the QC Report 121
To display visual results 121
3 Creating Grid Files 127
About Grid Templates and Grid Files 128
When to Use Manual Gridding 129
Analyzing the Initial Grid for an Agilent Image 130
10 Feature Extraction 12.2 User Guide
Contents
To open the current extraction in grid mode 130
If Automatic Gridding for an Agilent Image Fails 132
Step 1. Create a new project for failed extraction sets 132
Step 1b. Manually create a new project for failed extraction sets 133
Step 2. Access grid mode from a project extraction set 134
Step 3. Position grids and save grid file 134
Step 4. Set up and rerun Feature Extraction 135
Creating and Using a Grid File for non-Agilent Images Overview 136
Step 1. Start grid mode from the non-Agilent image file 136
Step 2. Position grids, locate spots and save grid file 137
Step 3. Extract the image file with its new grid file 138
Starting Grid Mode for a non-Agilent Image 139
To open an image file 139
To change the Portrait/Landscape view of the image 142
To open grid mode for a non-Agilent image 142
Viewing and Changing Grid Geometry 149
Optimizing Grid Placement Using the Help Wizard 151
To zoom in or out of the views 151
To undo or redo grid position movements 152
Step 1. Get nominal spot spacing manually 153
Step 2. Get nominal subgrid spacing manually 155
Step 3. Adjust the position of the main grid 156
Step 4. Adjust the subgrid 157
Step 5. Adjusting rotation and skew of subgrids 159
Step 6. Save grid file 159
Manual Grid Adjustments 160
To rotate the position of the main grid 160
To move or rotate the main grid with Grid Adjustment pane 161
To rotate the subgrid 162
To skew the subgrid corner 163
To position subgrids with Grid Adjustment pane 163
Feature Extraction 12.2 User Guide 11
Contents
Saving a Grid File 165
Calculating Spot Size and Spot Centroids 166
Locating Spot Centroids Manually 167
To find a spot to adjust 168
To position the spot centroid 170
To ignore spots 171
To undo or redo spot position movements 171
To position spot centroids with the Grid Adjustment pane 172
Refitting a Grid 174
Changing Default Settings for Grid Editing 176
Shortcuts to Help You Work with Grids and Find Spots 177
4 Changing Protocol Settings 181
Changing Protocol Settings 182
Protocol templates 182
To create a new protocol from a protocol template 183
Creating a Protocol From an Existing Protocol 184
To open a protocol 185
To view or change the protocol properties 186
To change the protocol steps to run 187
To change the settings for each protocol step 191
To save a protocol with a new name 192
Changing Protocol Step 1: Place Grid 193
To access the settings to place a grid 193
To select the microarray format 194
To select a method for placing the grid 196
To select to use enhanced gridding 196
To select to enable background peak shifting 197
To select to use central part of pack for slope and skew calculations 198
To select to use the correlation method to obtain origin x of subgrids 198
12 Feature Extraction 12.2 User Guide
Contents
Changing Protocol Step 2: Optimize Grid Fit 200
To access the settings to optimize the grid fit 200
To select the grid format 201
To change the settings to iteratively adjust corners when grid format is
selected 202
Changing Protocol Step 3: Find Spots 204
To access the settings to find spots when spot format is automatically
determined 204
To select the statistical method to reject pixel outliers 205
To select whether to use mean or median for data analysis 206
To enable the Enhance Spot Finding option (CGH protocols only) 206
To access the settings to find spots when spot format is selected 207
To change values to find and size the spots. 208
To select a spot statistics method to define features 209
To select to autoestimate the local radius to define local background 211
Final calculations for protocol step to find and measure spots 212
Propagation of Find Spot errors to other algorithms 212
Changing Protocol Step 4: Flag Outliers 213
To access the settings to flag outliers 213
To change settings to flag population outliers 214
To change settings to flag non-uniform outliers 215
Changing Protocol Step 5: Compute Bkgd, Bias and Error 220
To access the settings to compute background, bias and error 221
To select the background subtraction method 222
To choose to remove the surface trend found in the data 225
To choose to adjust the background globally (2-color) 227
To choose to perform multiplicative detrending 228
To choose to use robust negative control statistics 229
To select the method to estimate the error (Error Model) 229
To choose to use or not use surrogates 233
Propagation of errors from the Compute Bkgd, Bias and Error algorithm to other
algorithms 234
Feature Extraction 12.2 User Guide 13
Contents
To change settings to test the significance of a feature 234
Changing Protocol Step 6: Correct Dye Biases 237
To access the settings to correct dye biases 237
To choose how to select dye normalization list 238
To choose a method to select a set of dye normalization probes 239
To select to omit background population outliers 240
To select to allow positive and negative controls 240
To select signal characteristics 241
To select a normalization correction method 241
Changing Protocol Step 7: Compute Ratios 243
To access the settings to compute ratios 243
To enter the value for the log ratio limit 244
Use of surrogates for log ratio and p-value calculations 244
Changing Protocol Step 8: MicroRNA Analysis 248
To access the settings to analyze microRNA (miRNA) images 248
To choose to generate the results to a GeneView file 249
To choose to scale the analysis with the effective feature size
parameters 250
To change the Is Probe Detected multiplier 250
To choose to exclude non detected probes 251
Changing Protocol Step 9: Calculate Metrics 252
To access the settings to calculate metrics 252
To select to use spikein targets 253
To change the minimum population for replicate statistics 253
To select the grid test format 253
To change the p-value for differential expression 254
To change the percentile value 254
Changing Protocol Step10: Generate Results 255
To access the settings to generate results 255
To select the type of QC Report 255
To select to generate a single file for the text output 256
14 Feature Extraction 12.2 User Guide
Contents
To change the compression factor for the JPEG output 256
5 Changing Image Displays 257
Displaying and Saving Image Files 258
To open an image file 258
To crop an image 259
To save a cropped image to a new file 261
To add or change the New Array Identifier 263
Loading Feature Extraction Visual Results 264
To view visual results for low-resolution images 265
To view visual results for high-resolution images (2- by 3-micron or
20-bit) 266
To save an image (original or cropped) for import to another program 269
To display file information 270
Changing Image Display Settings 271
To change image orientation (landscape/portrait) 271
To change the channels that display 272
To change the display color 273
To change between log and linear color scale 276
To change the intensity scale (color display range CDR) 276
To change the magnification (zooming) 279
To change the window size to fit the image 281
To change the extraction results display 282
To change and save histograms and line graphs 283
To change default image viewing settings 287
To change the print settings 289
Hot key actions for image displays 291
Changing Feature Extraction Window Displays 293
To move windows and panes 295
To select the layout to use when program starts 296
Index 299
Feature Extraction 12.2 User Guide 15
16
Contents
Feature Extraction 12.2 User Guide
Agilent Feature Extraction 12.2 User Guide
1 Working with Feature Extraction
Feature Extraction Process 18
Overview of Feature Extraction Capabilities 21
What You Can Do to Set up and Run Automatic Batch Extraction 25
What You Can Do with Images Before Extraction 27
What You Can Do with Protocol Algorithms 29
What You Can Do with.shp Files of Extracted Images 33
What You Can Do with Extraction Results 34
This chapter gives you an introduction to the process of Feature Extraction and lets you know what you can do to optimize the process.
17Agilent Technologies
1 Working with Feature Extraction Feature Extraction Process
Feature Extraction Process
Feature Extraction consists of two
major processes: image analysis to
place the grid and locate spots and
data analysis to define and
measure spot features for gene
expression.
18
This section contains a flow chart that shows the process for producing feature extraction on Agilent or non- Agilent microarrays scanned on the Agilent Scanner. The data analysis flow starts with removing outlier pixels. It ends with transfer of Feature Extraction results into third- party data analysis programs to compare the results of different experiments. This flow chart shows the protocol steps for a Feature Extraction using a CGH protocol.
When you load an Agilent microarray file generated by an Agilent scanner into a project, the program automatically assigns a grid template, based on the array AMADID, and a protocol.
The protocol uses the grid template (array design) to place the grid and optimize the grid fit. It automatically locates the spots for Agilent microarrays.
To locate spots on non-Agilent microarray files generated by the Agilent scanner, interactively position the grid and save the grid file, which you load into the project.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 Feature Extraction Process
Feature Extraction 12.2 User Guide
After the program places the grid and finds spots, it uses the remaining Feature Extraction algorithm parameters contained in the Feature Extraction protocol to carry out data analysis.
Flags features and background regions that are population and non- uniformity outliers.
Computes background, bias, and error and performs corrections for these as specified in the protocol settings.
If the image is 2-color, it performs dye normalization and, for gene expression, calculates a p-value that is a confidence measure of gene differential expression.
Calculates the log ratios of the dye-normalized green and red channel signals, and calculates final error and significance value.
19
20
1 Working with Feature Extraction Feature Extraction Process
Generates the stats table and calculates QC metrics for the QC Report.
Generates for each extracted image the output files (Text, MAGE, QC Report, etc.) that you selected in the Project Properties.
Optionally, the program automatically sends the Feature Extraction results to certain third-party data analysis programs, such as Rosetta Resolver (via FTP).
You can also run a Feature Extraction workflow from within Agilent Genomic Workbench, and automatically analyze the extracted data.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 Overview of Feature Extraction Capabilities
Overview of Feature Extraction Capabilities
Feature Extraction program window
Feature Extraction 12.2 User Guide
Automatic batch extraction
You create projects to contain extraction sets. The extraction set includes the Agilent or non- Agilent image file, a grid template or grid file, used to locate spots, and a protocol to specify the values for the data analysis parameters.
A list of supported scans and array formats are listed in the following table. Non- Agilent scans are not supported.
Table 1 Supported scans and array formats
Array Format Scanner Scan type Resolution Support Project run summary message
Agilent 65um Agilent 16 bit 5um Supported and tested
Agilent 65um Agilent 20 bit 5um Supported and tested
Agilent 30um Agilent 16 bit 3um Supported and tested
Agilent 30um Agilent 20 bit 3um Supported and tested
Agilent 65um Agilent 16 bit 3um Supported but not
tested
Information message
Agilent 65um Agilent 16 bit 2um Supported but not
tested
Information message
Agilent 65um Agilent 20 bit 3um Supported but not
tested
Information message
Agilent 65um Agilent 20 bit 2um Supported but not
tested
Information message
Agilent 30um Agilent 16 bit 2um Supported but not
tested
Information message
Agilent 30um Agilent 20 bit 2um Supported but not
tested
Information message
21
1 Working with Feature Extraction Feature Extraction program window
Agilent 65um Agilent 16 bit 10um Not Supported Warning then proceed
Agilent 65um Agilent 20 bit 10um Not Supported Warning then proceed
Agilent 30um Agilent 16 bit 10um Not Supported Warning then proceed
Agilent 30um Agilent 16 bit 5um Not Supported Warning then proceed
Agilent 30um Agilent 20 bit 10um Not Supported Warning then proceed
Agilent 30um Agilent 20 bit 5um Not Supported Warning then proceed
non-Agilent N/A Agilent any any Limited Support Information message
Agilent 30um non-Agilent 2-color 2um Select Support Information message
Agilent 65um non-Agilent 2-color 5um Select Support Information message
Table 1 Supported scans and array formats (continued)
Array Format Scanner Scan type Resolution Support Project run summary message
22
Explanation of support types Not Supported means that no support is provided by Agilent Technologies. Limited Support means that Agilent provides support on a best- effort basis to help you scan microarrays and extract images, however, there is no guarantee or commitment. Select Support is provided for Agilent arrays scanned on selected non- Agilent scanners (specifically, the Roche NimbleGen MS 200 scanners, Molecular Devices GenePix scanners, select Innopsys InnoScan scanners, and select Tecan scanners). See the scanner compatibility matrix for more information.
Although Feature Extraction 12.2 is capable of processing images of
Devices, Innopsys, and Tecan scanners, Agilent can only guarantee the
quality of the data output when processing images of Agilent arrays
scanned on an Agilent scanner.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 Feature Extraction algorithms for data analysis
Feature Extraction 12.2 User Guide
When you start an extraction, the program automatically extracts all the sets in the project and produces a project summary report and a QC report for each image. You can also set up a project to extract Agilent microarray images automatically in real time, as soon as the files are made available by the Agilent scanner. For images produced on a NimbleGen, Molecular Devices, Innopsys, or Tecan scanner, the program first converts the image and then adds the image to the extraction set.
Image analysis
Before extraction, you can examine the spots on your microarrays visually and compute statistics about the image, using the histogram and line plots provided in the program. You can interactively position grids to find the spots on the microarray and calculate the spot centroids and sizes.
Feature Extraction can handle high- resolution images (2- or 3- micron and 20- bit) from Agilent scanners. To avoid using large amounts of memory to handle the high- resolution images, Feature Extraction loads a low- resolution preview image for first viewing.
When you create a cropped image to one- tenth of its original size or less, the program loads the original high- resolution image for the cropped image.
Histogram and line plots are always based on the original high- resolution image, not the preview one.
Feature Extraction algorithms for data analysis
For all images, the protocols define features and background regions, remove pixels, and flag features and background regions that affect the reliability of the results. In addition, protocols set up the extraction to subtract background information from features and make a background adjustment on signals of low intensity.
23
24
1 Working with Feature Extraction Extraction results
For GE 2- color, CGH, and ChIP (including DNA Methylation) images, the program normalizes for any differences in dye signal intensity. (SNP probes are not included in these calculations.) Then the program calculates a reliable log ratio, p- value, and log ratio error for each feature. These values give you a confidence measure in the measured log ratio.
For microRNA (miRNA) images, Feature Extraction uses a special analysis, described in more detail in Chapter 4.
Extraction results
Feature Extraction offers several types of outputs for each microarray extraction:
Project Run Summary report
MAGE (for Rosetta Resolver)
JPEG (for Rosetta Resolver)
Text (normal and .zip format)
Shape
Grid (used in Feature Extraction only)
QC Report
You can export the text results containing all the parameters, statistical calculations, and feature results into a spreadsheet.
To compare the results of different microarray experiments, you can also import various output types into GeneSpring, Agilent Genomic Workbench, and third- party analysis programs. The text results can be saved in full, compact, QC, and minimal formats, and MAGE- ML results can be saved in both full and compact formats. For details on what information is included in these files, see the Feature Extraction 12.2 Reference Guide.
You can load and display the results to see the outcome of the statistical and flagging algorithms for each feature and background.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 What You Can Do to Set up and Run Automatic Batch Extraction
What You Can Do to Set up and Run Automatic Batch Extraction
Setup
See the Feature Extraction Quick
Start Guide for definitions of
project, extraction set, grid
template, and protocol.
Feature Extraction 12.2 User Guide
You can perform the following tasks when you set up automatic batch extraction:
Create projects and save them with new names.
Set up the program to log on to the eArray server. eArray automatically downloads the required grid templates with associated dye normalization gene lists (if available), protocols, and associated QC metric sets that are missing in the database.
Load image files to create extraction sets. For Agilent image files, the program automatically:
Loads the required grid template (with associated DyeNorm List, if available) and associated default protocol for each extraction set, if available in the database.
Checks the eArray server for the required grid templates and protocols that are missing from the database.
Drag and drop scan files onto the project.
Type or change the protocol or grid template for the project.
Drag and drop grid templates or protocols onto each extraction set.
Assign a grid file to an extraction set.
Choose the type of result file to save.
Set up to transfer the files via FTP after extraction.
Delete extraction sets from the project.
Remove grid templates and protocols from the database.
Add grid templates, protocols, and QC metric sets to the database.
Edit protocols and grid template properties.
25
26
1 Working with Feature Extraction Run extraction
Change the order that extraction sets are processed.
Choose to stop Feature Extraction automatically after a specified time when extraction is in real- time mode (On- Time Project).
Load 2 micron, 3 micron, 5 micron, and 16- or 20- bit image files.
Create a critical probes list from within the Grid Template dialog box.
Run extraction
Start and stop extracting.
Monitor the extraction progress for each extraction.
View a Summary Report of the extraction set (or batch of extraction sets).
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 What You Can Do with Images Before Extraction
What You Can Do with Images Before Extraction
View image and change image display
For details on how to view and
change image display, see
Chapter 5, Changing Image
Displays.
Feature Extraction 12.2 User Guide
Change color and/or scale of the image for easier viewing.
Select rectangular and elliptical regions of interest.
Create histograms and line graphs.
Create images for other programs.
Print the images.
Compute statistics about the image.
Line plots
Histogram plots
Type and save barcode information for Agilent images where the barcode reader has failed.
Flip and rotate the image from landscape to portrait mode and back, for scans of third- party arrays (cannot do for high- resolution images).
Find grid and spots on the image manually
For details on manual gridding, see
Chapter 3, Creating Grid Files.
For Agilent images, grid templates are applied automatically. You normally do not need to adjust the grid, unless an error indicating a gridding failure appears in the Summary Report for an extraction. For non- Agilent images, create a grid file before you extract the image.
Interactively adjust the grids on the microarray scan.
Interactively adjust the spots on the microarray (for third- party microarray images only).
Calculate the spot size and compare the position of the nominal spot centroid (+) laid down by the grid with the found centroid position (X) for the spot (for third- party microarray images only).
27
28
1 Working with Feature Extraction Find grid and spots on the image manually
Move the found centroid position (X) to where you want it on the spot (for third- party microarray images only).
Feature Extraction uses these manually adjusted centroid positions to define features, rather than positions calculated by the Place Grid algorithm.
Find features by typing feature number, gene name, or probe name (for third- party microarray images only). For Agilent arrays, if a shape file is loaded, you can find features by feature number.
In Grid Mode, select spots to ignore from analysis (for third- party microarray images only).
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 What You Can Do with Protocol Algorithms
What You Can Do with Protocol Algorithms
For details on protocols, see
Chapter 4, Changing Protocol
Settings in this User Guide and
the Feature Extraction 12.2
Reference Guide.
Feature Extraction 12.2 User Guide
Protocols contain the values for the parameters used by Feature Extraction algorithms to calculate intermediate and final results. See Feature Extraction Process on page 18. When you install Feature Extraction, default protocols are loaded into the database one protocol for each microarray type (CGH, ChIP, GE1, GE2, and miRNA). The protocols are identified using the following naming convention:
What you can do with the protocol steps
Select the protocol steps that you want to use.
Change the default settings for each algorithm in the protocols.
Reload default settings after you have changed them.
What the protocol steps do for you
These protocol algorithms operate in the following order.
Place Grid
For any microarray loaded with a grid file that you saved during Grid Mode (during interactive gridding), automatically calculates centroid positions. Any spots that were manually adjusted are left unchanged.
29
30
1 Working with Feature Extraction What the protocol steps do for you
For any microarray, provides two methods of autogridding. The default method is to allow some distortion. This method lets the grid have a feature spacing somewhat different from the spacing found in the grid template. The second choice is place and rotate only. This method assumes that the feature spacing from the grid template is accurate and only lets this grid move to the best location.
For any microarray, you can set the Enable Background Peak Shift flag in order to get better gridding. Some variations in the results are expected with and without use of this flag as the grid positions obtained differ.
For any microarray, you can select to use regions near to the center of packs for slope and skew calculations. This option is of use in cases where pack edges have dim spots and are failing to grid. When this option is selected, you can select to have the program to perform one extra step of correlation following the projection data analysis to get the origin.
Optimize Grid Fit
Find Spots
Calculates statistics for the pixel populations of the features and backgrounds, removing outlier pixels.
Calculates signal statistics for inlier pixels.
Flag Outliers
Compute Bkgd, Bias and Error
If you select to subtract the background, subtracts background from the raw signal of each feature using a local or global background method that you select.
Corrects for spatial gradient across each microarray and any dome effects.
Adjusts the subtracted signal for inaccuracies in the noise level estimation.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 What the protocol steps do for you
Feature Extraction 12.2 User Guide
For all data, also calculates the error based on a selected error model.
Determines if the feature signal is greater and significantly different from background using the error model calculation.
Can calculate signals and background for low- intensity features using surrogates.
Correct Dye Biases
Automatically determines if a dye normalization gene list exists for the grid template.
Selects to use dye normalization list or generates one internally.
Selects the features for dye normalization based on a feature selection method of your choosing.
Performs dye normalization on the normalization feature set with a curve fit of your choosing. (SNP probes are not included in dye normalization calculations.)
Compute Ratios
Calculates a final error of the log ratio and a significance value (p- value) based on an error model that you selected in the Compute background step.
MicroRNA Analysis
Only available in the miRNA protocol.
Calculates the TotalGeneSignal for microRNA (miRNA) microarray images only.
This analysis includes all the previous Feature Extraction steps but does not correct for dye biases, or compute ratios.
See Chapter 4, Changing Protocol Settings of this guide to learn how to change protocol settings for the miRNA microarray image extraction.
See the Feature Extraction 12.2 Reference Guide to learn more about the algorithm in this step.
Calculate Metrics
Generates the stats table and calculates QC metrics for the QC Report.
31
32
1 Working with Feature Extraction What the protocol steps do for you
To generate a complete stats table and QC report, leave this option turned on. Some values are not included if this option is turned off.
See the Feature Extraction 12.2 Reference Guide for details on QC metrics.
Generate Results
Generates for each extracted image the output files (Text, MAGE, QC Report) that you selected in the Project Properties.
Also generates a file for the microRNA image, in addition to the output files (Text, MAGE, QC Report) that you selected in the Project Properties.
The file is a GeneView.txt file and contains the SystematicName, the ControlType, the gTotalGeneSignal and the gTotalGeneError for the miRNA microarray.
See the Feature Extraction 12.2 Reference Guide to learn more about this text file.
Feature Extraction 12.2 User Guide
Working with Feature Extraction 1 What You Can Do with.shp Files of Extracted Images
What You Can Do with.shp Files of Extracted Images
Feature Extraction 12.2 User Guide
Show and hide feature information on the image:
Feature number
Gene name
Raw intensities per color channel
Log ratio (rProcessedSignal/gProcessedSignal) for 2- color microarray images
Show or hide outlier features on the image.
Search and display one feature at a time using feature number.
Get reference to annotation results from Help > Reference Guide.
Print the images.
View the pixels used in the calculation of a feature or set of features.
33
1 Working with Feature Extraction What You Can Do with Extraction Results
What You Can Do with Extraction Results
34
View a QC report for GE 2- color, GE 1- color, and miRNA, each with or without Agilent spike- ins, and for CGH. You can use the QC report to evaluate if the grid was placed correctly (snapshot of four corners of the grid) The QC report also shows if the microarray scan data is reliable.
View all the extraction parameters, statistical calculations, and feature results created by the Feature Extraction program to know how reliable each feature result is.
Export the text file containing the results to a spreadsheet or a database of your choice.
NOTE Extracted data is not sent to Agilent QC Chart Tool automatically. To add
extracted data to QC Chart Tool, import the data using the Import > FE Stats and Parameters command.
Automatically export MAGE- ML files to Rosetta Resolver (FTP) or manually load the files into a database of your choice.
Feature Extraction 12.2 User Guide
Agilent Feature Extraction 12.2 User Guide
2 Extracting Microarrays Automatically
Setting up eArray Login for Automatic Updates 36
Setting Up and Editing Projects 39
Changing Properties for Individual Projects 45
Changing Default Project Properties in Preferences 57
Working With Extraction Sets 71
Working With Grid Templates/Files 78
How Dye Normalization Lists are Associated With Array Designs/Grid
Templates 80
Setting Up and Changing Grid Templates and Files 83
Working With Protocols 104
Working With QC Metric Sets 108
Running Feature Extraction 113
Displaying Results 115
You set up to extract a microarray image through the use of projects. Projects let you automatically extract one or more image files as either:
A batch of existing Agilent or non- Agilent files that you add to the project (Standard Project)
One Agilent image at a time as scanning completes (On- time Project)
This chapter explains how to set up and edit projects, run feature extraction, and view the results.
35Agilent Technologies
2 Extracting Microarrays Automatically Setting up eArray Login for Automatic Updates
Setting up eArray Login for Automatic Updates
36
Feature Extraction can connect to the Agilent eArray server to automatically download, install, and use the required grid templates, dye normalization gene lists, Feature Extraction protocols, and QC metric sets during extraction.
It is highly recommended that you set up an eArray login before you run the extraction project.
1 Double- click the Feature Extraction shortcut on your
This shortcut was created when you installed the Feature Extraction program.
When you first open the Feature Extraction program, the eArray Login Setting window appears on top of the Feature Extraction Standard Project main window. See Figure 1 on page 36. If you do not see the eArray Login Setting window, click Tools > eArray Settings.
Figure 1 eArray Login Setting dialog box
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Setting up eArray Login for Automatic Updates
Feature Extraction 12.2 User Guide
2 Type your User Name and Password in the eArray Login Setting window, if you already have an existing eArray account.
Hint: The user name is the e- mail address you provided when you created an eArray account.
If you do not have an eArray account, register for one by clicking the Register button. This opens the Agilent eArray registration website, where you can request a free eArray account. Follow the instructions provided on the eArray website to complete the registration.
3 Click Test Connection. The message Login Successful!!! pops up if the connection to the eArray server is successful. If you get the message Login denied, check that the user name and password you typed are correct.
4 In the Advanced Options, check the appropriate boxes:
Use eArray server during extraction
Select this box to let Feature Extraction log onto the eArray server to download, install, and use grid templates (array design files), dye normalization gene lists, Feature Extraction protocols, and QC metric sets during extraction.
Check for updates of grid template
Select this check box to download updates to existing grid templates already in the Grid Template Browser. Updated grid templates are added to the database, leaving existing grid templates, unless you mark Replace old grid template.
If more than one version of a grid template (design) is in the database, and you want to perform a feature extraction using an older grid template version, clear this check box. Otherwise, Feature Extraction automatically uses the newest version of the grid template (whether or not you selected to use the eArray server during extraction).
37
38
2 Extracting Microarrays Automatically Setting up eArray Login for Automatic Updates
Replace old grid template
When this check box is selected, Feature Extraction replaces old grid templates in the Grid Template Browser with the latest ones from the eArray server.
On starting Feature Extraction check for protocol update from eArray server
When this box is selected, Feature Extraction downloads updates to existing default protocols in the FE Protocol Browser from the eArray server every time you start the Feature Extraction program.
5 To save the eArray login settings, click Save.
You get a message Setting updated successfully. Click OK to close the message.
6 Click Close to exit the eArray Login Setting window.
Now you see the Feature Extraction Standard Project window with FE_Project1 displayed in the Project Explorer, as shown in Figure 2 on page 40.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Setting Up and Editing Projects
Setting Up and Editing Projects
Feature Extraction 12.2 User Guide
You set up to extract a microarray image through the use of projects. Projects let you automatically extract one or more image files as either:
A batch of existing Agilent or non- Agilent files that you add to the project (Standard Project)
One Agilent image at a time as scanning completes (On- time Project)
To set up or edit a standard project to run existing image files
You can run Feature Extraction for Agilent or non- Agilent microarrays automatically through the use of standard projects. To each standard project, you add one or more extraction sets, which consist of the image file, a grid template or file and a protocol. For images produced on a NimbleGen, Molecular Devices, or Innopsys scanner, the program first converts the images before adding them to the extraction set. The protocol contains all the values and settings for the parameters used in each step of the Feature Extraction process.
To learn more about working with individual images and with the Feature Extraction program window, see Chapter 5, Changing Image Displays.
39
40
2 Extracting Microarrays Automatically To set up or edit a standard project to run existing image files
Figure 2 Feature Extraction standard project window before adding extractions
If you do not see the QC Metric Set
Browser, click View > QC Metric Set Browser.
1 Create or open a Standard Project, if one is not already open.
See Create a Standard project on page 42.
2 If necessary, change Project Properties.
See To display project properties for individual projects on page 45.
3 Add the image files to make up extraction sets.
See Working With Extraction Sets on page 71.
4 View the Extraction Set Configuration tab sheet to make
See To display/edit the Extraction Set Configuration table on page 76.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To set up or edit an On-Time project to run image files in real time
Feature Extraction 12.2 User Guide
5 For any extraction set with a non- Agilent microarray image, add an appropriate grid file.
See To add or change a grid file to an extraction set on page 84.
6 For any extraction set with a non- Agilent microarray image, add a protocol.
See To add a protocol to an extraction set on page 105.
Starting with Feature Extraction 10.5, protocols are linked to QC metric sets. A QC metric set is no longer assigned in the Project Properties tab. Instead, a QC metric set is assigned to a protocol. You can add or change the default QC metric set associated to the protocol.
See Working With QC Metric Sets on page 108.
7 Save the project.
See To save a project on page 44.
To set up or edit an On-Time project to run image files in real time
You can run Feature Extraction for Agilent microarrays in real time as image files are saved by the Agilent Scanner. The differences between a Standard project and an On- Time project are listed below:
You can set up an On- Time project only for Agilent microarray images. Non- Agilent microarray images are not supported.
You do not add image files or extraction sets (image files, grid templates, and protocols) to the On- Time project.
You can type a maximum total processing time and a maximum time between extractions for an On- Time project.
For images produced on a NimbleGen, Molecular Devices, or Innopsys scanner, the program first converts the images before extracting them.
41
42
2 Extracting Microarrays Automatically To create a project
If you want the On- Time project to process microarrays whose grid templates are missing from the database, set up the program to use the eArray server during extraction. The required grid templates, protocols, and QC metric sets are then automatically downloaded and installed in the database. See Setting up eArray Login for Automatic Updates on page 36.
1 Open an On- Time project.
See Create an On- Time project on page 42.
2 Change Project Properties.
See To select a time to terminate real- time extractions on page 55.
3 Save the project.
See To save a project on page 44.
To create a project
Create a Standard project
To create a new standard project:
Click File > New > Standard Project, Or Click the New project file icon.
When you create the next project after FE_Project1, the name is FE_Project2. The name of every project that you create after that follows in numerical order, whether you save FE_ProjectN or not, or close FE_ProjectN or not.
Each time you restart the program, the first new project is FE_Project1, if that is the selection in the Preferences dialog box.
Create an On-Time project
Click File > New > On- Time Project.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To open a project
Feature Extraction 12.2 User Guide
Notice that the On- Time Project, Figure 3 does not have a Project Explorer and an Extraction Set Configuration tab.
Figure 3 On-Time Project Properties window
To open a project
To open an existing Standard or On- Time project:
1 Click File > Open > Project, Or Click the Open project file icon.
2 Select the project name, and click Open.
43
2 Extracting Microarrays Automatically To save a project
To save a project
44
To save a Standard or On- Time project to a new name:
1 Click File > Save As.
The default name is FE_ProjectN, where N is the number of the currently loaded project.
2 Type the new name, then click Save.
If you are saving the project for the first time, you can click the Save icon and type any name you choose. The default name that appears is FE_ProjectN, where N is the number of the currently loaded project.
The default folder for projects is C:\ProgramData\Agilent\
FeatureExtractionStandalone\FEProjects.
To save a project to its current name:
Click the Save icon.
If the current project has never been saved, the Save As dialog box appears.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Changing Properties for Individual Projects
Changing Properties for Individual Projects
If you want to change a property
for all projects, change the default
properties in the Preferences
dialog box. See Changing Default
Project Properties in
Preferences on page 57.
Feature Extraction 12.2 User Guide
When you create a new project or open an existing project in the Feature Extraction program, you see a Project Properties tab. The default values found on this tab were set in the Preferences dialog box. For each project that you open, you can change these default properties.
To display project properties for individual projects
Project Properties tab for Standard projects
In the main window for Project setup, click the Project Properties tab.
The Project Properties tab for Standard projects with specified extraction sets looks like Figure 4 below.
45
46
2 Extracting Microarrays Automatically To display project properties for individual projects
Figure 4 Standard Project Properties tab, expanded
Project Properties tab for On-Time projects
When you create a new On- Time project or open an existing one, you see the Project Properties tab.
The Project Properties tab for On- Time projects looks like Figure 5 below.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display project properties for individual projects
Feature Extraction 12.2 User Guide
Figure 5 On-Time Project Properties tab, expanded
The parameters and settings in this tab are the same as those in a standard project with these differences:
No capacity to select an available grid file.
No setting for default protocol.
No capacity to select an external DyeNorm list file.
Capacity to set up a folder for images coming in from the scanner.
See To select the incoming image folder for On- Time projects on page 55.
47
48
2 Extracting Microarrays Automatically To change general project property settings
Capacity to terminate the run automatically.
See To select a time to terminate real- time extractions on page 55.
Capacity to automatically extract XDR (extended dynamic range) images.
See To turn off the option to automatically extract XDR images on page 56.
To change general project property settings
Change to new values
Double- click the cell to the right of the field name.
True changes to False and vice versa. For list selections, each double- click brings up a different choice in the list.
This does not work for check boxes or Browser cells.
Or, click the arrow to the right of the cell, and select from the list.
Reset changed values to default values
At any time, you can change a modified value back to the original default value.
Right- click the cell on the right, and click Reset to Default.
Change the name of the operator
The only general property that you can change is the name of the Operator.
Click the cell next to Operator, and type the name of the operator.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the location for saving and/or sending output files
Feature Extraction 12.2 User Guide
View the Number of Extraction Sets Included
You cannot change this field. If you add more extraction sets than are visible on screen, you can refer to this field to check the number that you have added.
To select the location for saving and/or sending output files
See the Feature Extraction 12.2
Reference Guide for a listing of all
the results that are generated in
the output files.
This portion of the Project Properties tab lets you select the ultimate locations for the image and result files. See Outputs on page 49.
Outputs
Click the arrow to the right of the cell next to each of the result or image file types to select one of the following locations.
Local file only
FTP send only
Both local file and FTP send
Select to save the files both to a local file folder and to an external location.
The following output files are generated and sent to the locations you specify:
MAGE result file
This file contains results in the MAGE- ML (XML) data transfer format, which lets you transfer the data into Rosetta Resolver and Array Express databases.
You can select to send a Full Output Package of MAGE- ML data or a Compact (reduced) set of data. You can also choose to send the data in a Compressed (ZIP) format, or not.
If MAGE output is selected, you also have the capacity to set and pass data ownership and permissions to Rosetta Resolver.
49
50
2 Extracting Microarrays Automatically To select the location for saving and/or sending output files
Expand the MAGE folder, and expand the Rosetta Resolver Data Access Setting folder to view the following fields:
If you do not type names for these
settings, no access control
information is entered into the
MAGE-ML file.
Rosetta Resolver Data Access Settings: User Name Type the name of the user who owns the data transferred to the Rosetta Resolver database.
Rosetta Resolver Data Access Settings: Control Groups Type the names of the groups who have permission to use the data transferred to the Rosetta Resolver database. If there is more than one group, separate their names with a colon. For example, if you want the groups Public and Special to see the incoming extraction information, type public:special.
JPEG image file
You can control the spatial compression of this JPEG output in the protocol. See To change the compression factor for the JPEG output on page 256 of this guide.
Text result file
Compact mode contains all the
columns needed for loading into
GeneSpring, DNA Analytics, and
Agilent Genomic Workbench.
You can split this file into three files, each containing one of the sections. You can also import this file into GeneSpring data analysis program, either as a Full set of data, a Compact (reduced), QC, or Minimal. For a listing of the columns included in these formats, see Chapter 3 of the Feature Extraction Reference Guide.
You can select to save the text result file in .zip format. The default for this selection is False.
Visual Results
Grid
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the location for saving and/or sending output files
QC Report
Feature Extraction 12.2 User Guide
A QC Report is saved to the results directory as a pdf file. The type of QC Report saved is determined by the protocol type for the extracted data: 2- color gene expression (GE), 1- color GE, CGH (full or streamlined), ChIP, or miRNA. For some of the applications (GE1, GE2, and miRNA), spike- in data is also included in the QC Reports.
FTP Send Tiff File
Because Tiff files already are located in a local file folder, you can select to send them to an external location via FTP, or not.
Double- click the cell to the right of FTP Send Tiff File, and select True or False.
Crop multipack image to
individual arrays
When this option is set to True, the multipack image is cropped after extraction is completed and stored as individual images. It is saved in the output folder, along with the other results. The original multipack image is retained.
Local File Folder
Lets you select the local file folder for saved results.
Select to save results in the same local folder as the image This option saves the result files in the folder that contains the images in the extraction sets.
Under Local File Folder, double- click the cell next to Same As Image to select True, if necessary.
Select to save results in a local results folder You can select any folder to save the results.
1 Double- click the cell next to Same As Image to select False.
2 Click the cell next to Results Folder.
3 Click the Browse button.
4 Select the folder to contain the results.
5 Click OK.
51
2 Extracting Microarrays Automatically To change the FTP settings to send result files automatically to Rosetta Resolver
To change the FTP settings to send result files automatically to Rosetta Resolver
52
If you use the same settings repeatedly, the values specific to your installation can be set up in the Preferences dialog box. See To change the default FTP settings for automatic data transfer on page 64.
In the Project Properties tab, expand the FTP Setting folder, and type the values for these FTP settings:
Machine Destination
This field is specific to your Rosetta Resolver server name. Please contact your system administrator for the name to use.
Port
User name
Password
File Destination Folder
Scan information is sent to this folder. For MAGE results, the default setting is mage (/import/mage on the server).
Confirm with your system administrator that the FTP account uses the import directory of the server as the default directory for the transfer.
To set up to automatically assign protocols
If you configured your eArray settings to use the eArray server (see Setting up eArray Login for Automatic Updates on page 36) when you run Feature Extraction, the protocol is assigned based on the application type specified in the grid template. The protocol selected always contains
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To set up to automatically assign protocols
Feature Extraction 12.2 User Guide
the application type name and have the latest date. Application types are: CGH, ChIP, miRNA, and Expression (GE1 and GE2).
Select the highest priority location for the default protocol
The program uses the protocol specified at this parameter step to complete an extraction set if a protocol is available in both the grid template and project. For more information on extraction sets, grid templates and protocols, see Working With Extraction Sets on page 71.
1 Click the arrow to the right of the cell next to Highest Priority Default Protocol.
2 Select Grid Template Default or Project Default.
Grid Template Default
If selected, the protocol specified in the grid template is assigned to the extraction set. If the grid template contains no protocol, then the program assigns the project default protocol. Use this selection if you add extraction sets to your projects that require different protocols.
Project Default
Change the project default protocol
1 Click the arrow to the right of the cell next to Project Default Protocol.
2 Select a protocol from the list.
If the protocol you want is not shown on the list, import the protocol into the database. See To import or export a protocol to or from the database on page 107.
53
2 Extracting Microarrays Automatically To select to use a grid file
To select to use a grid file
54
Double- click on the cell next to Use Grid file if available and select True or False from the list.
True
False
To select an external DyeNorm list file
To learn how to create an external
DyeNorm list from the
DyeNormList Editor, see To create
a new internal dye normalization
gene list for a grid template on
page 93.
If the project calls for a DyeNorm list and the image is a non- Agilent array, use an external DyeNorm list. This list has the column names in the following order: Probe Name, Gene Name, and Systematic Name. This is the same list order for a DyeNorm list exported by the DyeNormList Editor.
External DyeNorm lists are only
used with non-Agilent microarrays.
For Agilent microarrays, specify the
dye normalization list in the grid
template. See To change the dye
normalization gene list in the grid
template on page 92.
1 Click the cell next to External DyeNorm List File.
2 Click the Browse button.
3 Select the .txt file, and click Open.
To select to overwrite previous results
Double- click on the cell next to Overwrite Previous Results and select True or False.
True
False
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the incoming image folder for On-Time projects
To select the incoming image folder for On-Time projects
Feature Extraction 12.2 User Guide
This selection is for On- Time projects only.
1 To assign a folder for incoming images, click the Browse button in the cell to the right of Incoming Image Folder.
2 Select a folder for the incoming images, and click OK.
To select a time to terminate real-time extractions
This task is for On- Time projects only.
1 Expand Project Run Terminating.
2 Set one or both of the following settings to True or False.
Use Waiting Idle Time Out
If True, you can set an Idle Time in hours and/or minutes. The Idle Time is the time between the arrival of one scan file and the arrival of the next scan file into the real- time image folder.
Use Run Duration Time Out
If True, you can set a Duration Time in days, hours and/or minutes. The Duration Time is the time you can wait for the project Feature Extraction to complete.
3 If you set either or both settings to True, expand the Idle Time or Duration Time folder, if not expanded.
4 Type the times.
55
2 Extracting Microarrays Automatically To turn off the option to automatically extract XDR images
To turn off the option to automatically extract XDR images
56
This task is for On- Time projects only.
Double- click the cell to the right of Automatic XDR Extraction to change the default setting from True to False.
If the setting is True, the paired XDR images from the Agilent scanner are extracted in a combined fashion and a single output file set covering a broader range of data is generated.
If the setting False, the two images are extracted separately and two output file sets are generated.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Changing Default Project Properties in Preferences
Changing Default Project Properties in Preferences
Feature Extraction 12.2 User Guide
You set the default project properties in the Preferences dialog box. The defaults that you type or change here appear in the Project Properties tab when you create a new project.
To open the Preferences dialog box
If you want to change the default settings and values that appear in the Project Properties sheet for all projects, you open the Preferences dialog box.
1 Click Tools > Preferences.
The Preferences dialog box opens to the folder that was last opened.
Figure 6 Preference Settings window
57
2 Extracting Microarrays Automatically To display the General default project properties
To display the General default project properties
58
1 Open the Preferences dialog box. See To open the Preferences dialog box on page 57.
2 Open the FE Project folder.
3 Select General.
Figure 7 Preference SettingsFE Project General
To select to save the Project Run Summary Report
The Project Run Summary Report contains a history of the extractions run for a specific project. The Project Run Summary Report is initially saved to a temporary .rtf file (YourComputerName_LastBatchReport.rtf) in the directory containing the project files.
This file is overwritten every time an extraction is run on a project and contains only the latest summary report. If you want to save the summary report for each project for later viewing, specify that intention here.
1 Display the General default project properties. See To display the General default project properties on page 58.
2 Click the cell next to Auto save extracting report.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select to overwrite previous results
Feature Extraction 12.2 User Guide
3 Select from the list.
Always
Never
Prompt
To select to overwrite previous results
1 Display the General default project properties. See To display the General default project properties on page 58.
2 Double- click on the cell next to Overwrite Previous Results to select True or False.
True
False
To select a default folder for a new project
This option defines the initial folder into which you save a new project in the Save As dialog box.
1 Display the General default project properties. See To display the General default project properties on page 58.
2 Click on the cell next to Default Folder for New Project.
59
60
2 Extracting Microarrays Automatically To display the default project properties for Output and Data Transfer
3 Select one of these options from the list.
Last accessed project folder
Folder you accessed to save the previous project.
FE Default Project Folder
Folder set by the application. If you let the installer choose the default installation folder, the path of this folder is C:\ProgramData\Agilent\FeatureExtractionStandalone\FEProjects
To display the default project properties for Output and Data Transfer
1 Open the Preferences dialog box. See To open the Preferences dialog box on page 57.
2 Open the FE Project folder.
3 Select Output & Data Transfer.
Figure 8 Preference SettingsOutput & Data Transfer
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the default location for saving and/or sending output files
To select the default location for saving and/or sending output files
Feature Extraction 12.2 User Guide
Display the default project properties. See To display the default project properties for Output and Data Transfer on page 60
See the Feature Extraction 12.2
Reference Guide for a listing of all
the results that are generated in
the output files.
This portion of the Preferences dialog box lets you select the default locations for the image and result files. See Select the default location for output files on page 61 and Select the default local file folder for saved results on page 63.
Select the default location for output files
Click the arrow to the right of the cell next to each of the result or image file types to select one of the following locations.
Local file only
FTP send only
Both local file and FTP send
Select to save the files both to a local file folder and to an external location.
The following output files are generated and sent to the locations you specify:
MAGE result file
This file contains results in the MAGE- ML (XML) data transfer format, which lets you transfer the data into Rosetta Resolver and Array Express databases.
You can select to send a Full Output Package of MAGE- ML data or a Compact, or reduced, set of data. And you can choose to send the data in a Compressed (ZIP) format, or not.
If MAGE output is selected, you also have the capacity to set and pass data ownership and permissions to Rosetta Resolver.
61
62
2 Extracting Microarrays Automatically To select the default location for saving and/or sending output files
Expand the MAGE folder, and expand the Rosetta Resolver Data Access Setting folder to view the following fields:
If you do not type names for these
settings, no access control
information is entered into the
MAGE-ML file.
Rosetta Resolver Data Access Settings: User Name Type the name of the user who owns the data transferred to the Rosetta Resolver database.
Rosetta Resolver Data Access Settings: Control Groups Type the names of the groups who have permission to use the data transferred to the Rosetta Resolver database. If there is more than one group, separate their names with a colon. For example, if you want the groups Public and Special to see the incoming extraction information, type public:special.
JPEG image file
You can control the spatial compression of this JPEG output in the protocol. See To change the compression factor for the JPEG output on page 256 of this guide.
Text result file
A Compact data set contains
sufficient columns for use by
Rosetta Resolver, GeneSpring GX,
CGH Analytics or ChIP Analytics
programs. See the Feature
Extraction 12.2 Reference Guide for
a list of these columns.
You can split this file into three files, each containing one of the sections. You can also import this file into the GeneSpring data analysis program, either as a Full set of data, Compact (reduced), QC, or Minimal set. For details on what is included in these formats, see Chapter 3 of the Feature Extraction Reference Guide.
You can also choose to create a .zip file containing the text results.
Visual Results
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the default location for saving and/or sending output files
Grid
Feature Extraction 12.2 User Guide
Choose to save the grid information to grid.csvfiles (for Agilent and non- Agilent grids) and feat.csv files (for non- Agilent grids). It contains the output of all of the grid placement steps (e.g., the spot centroids and spot sizes).
QC Report
Crop multipack image to
individual arrays
When this option is set to True, the multipack image is cropped after extraction is completed and stored as individual images. It is saved in the output folder, along with the other results. The original multipack image is retained.
Send a Tiff file via FTP
Because Tiff files already are located in a local file folder, you can select to send them to an external location via FTP, or not.
Double- click the cell to the right of FTP Send Tiff File, and select True or False.
Select the default local file folder for saved results
Select to save results in the local image folder This folder is the folder that contains the images in the extraction sets.
Under Local File Folder, double- click the cell next to Same As Image to select True, if necessary.
Select to save results in a local results folder This folder can be any folder that you choose.
1 Double- click the cell next to Same As Image and select False.
2 Click the cell next to Results Folder.
3 Click the Browse button.
4 Select the folder to contain the results.
5 Click OK.
63
2 Extracting Microarrays Automatically To change the default FTP settings for automatic data transfer
To change the default FTP settings for automatic data transfer
64
1 Display the default project properties for Output and Data Transfer. See To display the default project properties for Output and Data Transfer on page 60.
2 In the Project Properties tab, expand the FTP Setting folder, and type the values for these FTP settings:
Machine Destination
This field is specific to your Rosetta Resolver server name. Please contact your system administrator for the name to use.
Port
User name
Password
File Destination Folder
Scan information is sent to this folder. For MAGE results, the default setting is mage (/import/mage on the server).
Confirm with your system administrator that the FTP account uses the import directory of the server as the default directory for the transfer.
To change default run post processing plug-ins
1 Display the default project properties for Output and Data Transfer. See To display the default project properties for Output and Data Transfer on page 60.
2 In the Project Properties tab, expand the Post Processing Plug Ins folder, and type or select the plug- in you want to run.
The plug- ins are executed independently, so you can run the same or different plug- ins after each event.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display the default properties for standard projects
Run After Each Array
Feature Extraction 12.2 User Guide
Select a plug in to run after each array is extracted. All array related arguments are passed to the executable called.
Run After Each Slide
Select a plug in to run after each slide is extracted, where a slide contains 1 or more arrays. Arguments pointing to files relating to the slide are passed to the executable called.
Run After Each Batch
Select a plug in to run after each batch of extractions is completed. Arguments pointing to files relating to the batch are passed to the executable called.
To display the default properties for standard projects
1 Open the Preferences dialog box. See To open the Preferences dialog box on page 57.
2 Open the FE Project folder.
3 Click Standard Setting.
Figure 9 Preference Settings Standard Project
65
2 Extracting Microarrays Automatically To change the default protocol and its location
To change the default protocol and its location
66
Select the default highest priority location for the default protocol
The program uses the protocol assigned to this location to complete an extraction set if a protocol is available in both the grid template and project. To learn more about extraction sets, grid templates and protocols, see Working With Extraction Sets on page 71.
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Click the arrow to the right of the cell next to Highest Priority Default Protocol.
3 Select Grid Template Default or Project Default.
Grid Template Default
If selected, the protocol specified in the grid template is assigned to the extraction set. If the grid template contains no protocol, then the program assigns the project default protocol. Use this selection if you add extraction sets to your projects that require different protocols.
Project Default
Change the default protocol for all projects
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Click the cell next to Project Default Protocol.
3 Select a protocol from the list.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select to use a grid file
To select to use a grid file
Feature Extraction 12.2 User Guide
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Double- click on the cell next to Use Grid file if available to select True or False from the list.
True
False
To change the Start Run Mode setting
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Double- click the cell next to Using Last Saved Layout to select True or False from the list.
The default setting is False. If True is selected, the program saves the layout changes (i.e. placement of panels and browsers) for the Standard Project and reloads this layout the next time you open Feature Extraction.
To select an external DyeNorm list file
External DyeNorm lists are only
used with non-Agilent microarrays.
For Agilent microarrays, specify the
dye normalization list in the grid
template. See To change the dye
normalization gene list in the grid
template on page 92.
If the project calls for a dye normalization list and the image is from a non- Agilent array, use an external DyeNorm list. This list has the column names in the following order: Probe Name, Gene Name, and Systematic Name. This is the same list order for a DyeNorm list exported by the DyeNormList Editor.
67
2 Extracting Microarrays Automatically To select to create XDR extraction sets automatically
To learn how to create an external
DyeNorm list from a DyeNormList
Editor, see To create a new
internal dye normalization gene list
for a grid template on page 93.
68
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Click the cell next to External DyeNorm List File.
3 Select the external dye normalization list .txt file, and click Open.
To select to create XDR extraction sets automatically
The default for this setting is True. When set to True, the program automatically creates one extraction set for the two images (_H and _L) scanned with extended dynamic range (XDR) turned on in the scanner.
1 Display the default properties for standard projects. See To display the default properties for standard projects on page 65.
2 Double- click the cell to the right of Automatically Create XDR Extraction Settoselect either True or False.
If you want to be able to add just one of the two images to the extraction set and have the second image loaded automatically, follow step 3.
3 Double- click the cell to the right of Search XDR image file pair on name matching onlytoselect either True or False.
True is the default selection. The pair images must be in the same folder for the program to load the second image automatically to the extraction set.
To display the on-time default properties
1 Open the Preferences dialog box. See To open the Preferences dialog box on page 57.
2 Open the FE Project folder.
3 Click OnTime Setting.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To select the default image folder
Feature Extraction 12.2 User Guide
Figure 10 Preference Settings On-Time Project Properties
To select the default image folder
1 Display the on- time default properties. See To display the on- time default properties on page 68.
2 Click on the cell next to Incoming Image Folder.
3 Click Browse.
4 Double- click the image folder, and click OK.
To select a default time to terminate real-time Feature Extraction
1 Display the on- time default properties. See To display the on- time default properties on page 68.
2 Expand Project Run Terminating.
3 Set one or both of the following settings to True or False.
Use Waiting Idle Time Out
If True, you can set an Idle Time in hours and/or minutes. The Idle Time is the time between the arrival of one scan file and the arrival of the next scan file into the real- time image folder.
69
2 Extracting Microarrays Automatically To select to automatically extract an XDR image pair
Use Run Duration Time Out
70
If True, you can set a Duration Time in days, hours and/or minutes. The Duration Time is the time you can wait for the project Feature Extraction to complete.
4 If you set either or both settings to True, expand the Idle Time or Duration Time folder, if not expanded.
5 Type the times.
To select to automatically extract an XDR image pair
The default setting is True. That is, the program automatically extracts both the High and Low signal images as a single extraction set when the XDR option is turned on in the scanner.
1 Display the on- time default properties. See To display the on- time default properties on page 68.
2 Double- click the cell to the right of Automatic Extraction of XDR images to change from True to False, or vice versa.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Working With Extraction Sets
Working With Extraction Sets
What is an extraction set?
Feature Extraction 12.2 User Guide
To each standard project you add one or more extraction sets, which consist of the image file, a grid template or file and a protocol. The protocol contains the values and settings for the parameters used in each step of the Feature Extraction process.
Figure 11 Extraction sets in Project Explorer
Extraction set
Image file
Grid template
Protocol
Grid file
extraction set
20-bit single scans replace XDR
dual scans. To determine if a scan
is 20-bit, click the image file info
button, click the Scan Statistics
tab, and view the Saturation Value
(about 777,000 for 20-bit).
Figure 12 XDR extraction image files marked with _H and _L
71
2 Extracting Microarrays Automatically How does the program create an extraction set?
grid template
72
Grid information from Agilent array design files stored in the database. A grid template includes feature annotation and general geometry about the microarray (number of rows, columns, subgrids, feature spacings), which is used to locate spots before data analysis takes place. Although not specific to any image, grid template information is usually applied to the image for which the template was designed.
grid file
protocol
How does the program create an extraction set?
Step a
AMADID stands for Agilent
MicroArray Design IDentifier.
The first 6 digits of the array design
file indicate the AMADID. Given a
design file 012097_D_20070820,
012097 is AMADID and 20070820
is YYYYMMDD that the design file
was generated.
For most Agilent microarray images, the grid template is automatically assigned, if the grid template is in the database. You may need to assign a grid template/grid file manually if, for example, the Agilent microarray image failed auto- gridding on a previous run.
Use a grid file for a non- Agilent extraction set, or if you had to manually grid the microarray.
If the grid template is not in the database, and you have selected to use the eArray server during extraction, the program gets the grid template for the array, based on its barcode, and uses that grid for the extraction. You can also
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To add extraction sets (image files) to the standard project
Feature Extraction 12.2 User Guide
select to have eArray check for updates to the grid template before extracting. See Setting up eArray Login for Automatic Updates on page 36.
Step b
The added protocol is either the default protocol entered with the grid template or specified for the project. You can select which protocol is the first chosen by the program, the protocol in the grid template or the one in the project. If the added grid template or file does not contain a protocol, the program uses the default protocol for the project.
To add extraction sets (image files) to the standard project
You can add an image file or extraction set in one of four ways:
Drag and drop the image file onto Project Explorer
Select the extraction set through the FE Project toolbar.
Select the image file through the Edit menu.
Select the image file through the Project Explorer shortcut menu
Note that for images produced on a NimbleGen, Molecular Devices, or Innopsys scanner, Feature Extraction converts the image when you start the extraction.
To add image files through Windows Explorer:
1 Open Windows Explorer.
2 Select the file or files that you intend to add.
3 Drag them to the Project Explorer and then drop them.
You now see a list of extraction sets.
4 Expand the extraction sets.
73
74
2 Extracting Microarrays Automatically To add extraction sets (image files) to the standard project
Figure 13 Project window after adding image files (extraction sets)
To add extraction sets through the toolbar:
1 Click the Add Extraction Sets icon.
2 Select the image files you want to add, then click Open.
You can select one or more images for extraction. Use the Shift key to select contiguous files or the Ctrl key to select separate files.
To add image files through the Edit menu:
1 Click Edit > Add Extraction Sets.
2 Select the image files you want to add, then click Open.
To add image files through the Project Explorer shortcut menu:
1 Right- click in Project Explorer.
2 Click Add Extraction.
3 Select the image files you want to add, then click Open.
You can also add extended dynamic range (XDR) extraction sets. These consist of a pair of images, one of high signal and the other of low signal, scanned with the XDR option turned on in the scanner. Feature Extraction extracts both images as a set and creates a single combined set of output files with data covering the combined dynamic range of both scans before it moves on to the next extraction set.
Make sure that the preferences options for creating and extracting an XDR extraction set are turned on. The default selections are True (ON).
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To sort extraction sets
To sort extraction sets
Feature Extraction 12.2 User Guide
If you want to run extraction sets in an order different from their order in Project Explorer, sort them using these instructions or drag and drop the extraction set in any order in Project Explorer.
1 Click Edit > Sort Extraction Sets.
2 Then select from the list By Name, By Image, By Grid or By Protocol.
The list is sorted alphabetically and numerically for any of the choices.
To remove an extraction set from the project
To delete an extraction set from Project Explorer
1 Select an extraction set in Project Explorer.
You can select only one extraction set in Project Explorer.
2 Right- click Project Explorer, and click Delete.
The extraction set is immediately deleted.
Or
Click the Delete Extraction Sets icon, or click Edit > Delete Extraction Sets.
If an extraction set in Project Explorer is selected, a message appears to ask if you want to delete the single set in Project Explorer. Click Yes.
If an extraction set or sets in the Extraction Set Configuration table is selected, a message appears to ask if you want to delete the extraction sets in the table.
75
76
2 Extracting Microarrays Automatically To display/edit the Extraction Set Configuration table
To delete an extraction set or sets from the Extraction Set Configuration table
1 Select an extraction set or sets in the Extraction Set Configuration table.
You can select more than one by holding down the Ctrl key while you click additional extraction sets.
2 Right- click the selected extraction sets, and click Delete Extraction sets.
The extraction sets are immediately deleted
Or
Click the Delete Extraction Sets icon, or click Edit > Delete Extraction sets.
If an extraction set or sets in the Extraction Set Configuration table are selected, a message appears to ask if you want to delete the extraction sets in the table. Click Yes.
If different extraction sets are selected in both Project Explorer and the Extraction Set Configuration table, a message appears to ask if you want to delete the extraction set in Project Explorer. Click Yes. To delete the selected sets in the Extraction Set Configuration table, repeat step 2.
To display/edit the Extraction Set Configuration table
1 Click the Extraction Set Configuration tab.
2 Drag the column marker on the header to expand each cell.
Figure 14 Extraction Set Configuration table
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display/edit the Extraction Set Configuration table
Feature Extraction 12.2 User Guide
The Extraction Set Configuration tab is a table that lets you review your extraction sets all in one place.
Extraction Set Name
The name of the extraction set. This name is automatically set to be the same as that of the scan file.
Grid Name
Click the cell and click the arrow to add or change this template or file from this column.
Double- click the cell to view the Grid Template Properties dialog box.
Protocol Name
Click the cell and click the arrow to add or change the protocol from this column.
Double- click the cell to view the Protocol Editor.
Output Name
Scan File Name
Double- click this name to see the image in the scan file.
Scan File Path
77
2 Extracting Microarrays Automatically Working With Grid Templates/Files
Working With Grid Templates/Files
78
A microarray design is the layout of probes on a single glass slide with a particular format. Each array design contains the biological and control probe groups, as well as a required Agilent quality control grid (grid template). When working in Feature Extraction, the Grid Template Browser displays the array designs from which the grid templates are derived. Designs that are displayed in the Grid Template Browser were imported manually from a hard disk location, or downloaded manually or automatically from the Agilent eArray server.
To manage grid templates
There are 3 ways for you to set up/obtain a grid template (from an array design):
Automatically find Agilent catalog array designs or your custom array designs from the eArray website under your workgroup account
When you set up an extraction, Feature Extraction first checks if the array design file for this AMADID (Agilent MicroArray Design IDentifier) is in the Feature Extraction database. If you or another user previously downloaded or worked with this array design, the design file may already be present.
If you selected in your eArray settings to Use eArray server during extraction, Feature Extraction checks on the eArray website for any recent array design file update for this AMADID. See To automatically update required and existing array designs/grid templates on page 86. If a newer version of the design file exists on the eArray website, or the design file is not present in the database, Feature Extraction automatically downloads the design file and uses it for extraction.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To manage grid templates
Feature Extraction 12.2 User Guide
Use Online Update in the Grid Template Browser to manually search for and import or update array design files from the eArray server.
Add the array design/grid template file manually
Array designs that cannot be found on the eArray website as a catalog design or under your workgroup account (for example, designs you have through private collaborations, or early- access Agilent catalog designs) are imported manually. See To add or remove a grid template on page 84.
After a grid template/array design is added manually, the grid template is displayed in the Grid Template Browser. Routine annotation update checks are not performed on these designs because they are not present on the eArray website.
79
2 Extracting Microarrays Automatically How Dye Normalization Lists are Associated With Array Designs/Grid Templates
How Dye Normalization Lists are Associated With Array Designs/Grid Templates
80
Array designs created on eArray.com have optional dye normalization list files associated with them. When Feature Extraction downloads or updates an array design (grid template), if an associated dye normalization list exists for that design, it is automatically imported and saved with the design (grid template) in the database. Once an array design is created in eArray, the dye norm list is fixed. Annotation updates can happen, yet the dye norm list of probes never change.
If you add an array design/grid template file manually, Feature Extraction does not automatically associate a dye normalization gene list file with the design. Associate the dye normalization gene list with the array design/grid template using the Grid Template Properties. See To change the dye normalization gene list in the grid template on page 92.
The following table shows how rules are applied when Feature Extraction updates array designs (grid templates) that have dye normalization gene lists. This table assumes that Use eArray server during extraction and Check for updates of grid templates options are selected in Tools > eArray Settings.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 How Dye Normalization Lists are Associated With Array Designs/Grid Templates
Table 2 Dye normalization list rules
Grid template/design has dye normalization list association?
Protocol use dye norm list selection
Action Result
Grid template not in database-
No
Automatically Determine Run a feature extraction Grid template/design is
automatically downloaded
from eArray and used for
extraction.
Dye normalization probe
selection method set in
the Correct Dye Biases
step of the protocol is
used.
Grid template not in database
- Yes
Automatically Determine Run a feature extraction Grid template/design is
automatically downloaded
from eArray, with
associated dye
norrmalization list, and
used for extraction.
Old version - Yes
New version - No
Automatically Determine Run a feature extraction The dye normalization
gene list from the older
version is automatically
associated with the new
grid template version,
which is then used for
extraction.
Old version - No
New version - Yes
Automatically Determine Run a feature extraction The dye normalization list
and the new grid template
version are used.
Old version - No
New version - No
Automatically Determine Run a feature extraction Feature Extraction uses
the dye normalization
probe selection method
set in the Correct Dye
Biases step of the
protocol.
Feature Extraction 12.2 User Guide 81
2 Extracting Microarrays Automatically How Dye Normalization Lists are Associated With Array Designs/Grid Templates
New version - Yes Automatically Determine Create a new dye
normalization list and
associate it in the grid
template properties
Run a feature extraction
The new dye
normalization list is used
for extraction, however,
the original dye
normalization list is still
available.
New version - Yes True Run a feature extraction The associated dye
normalization list
associated with the
design is used.
Old Version - No
New Version - No
True Run a feature extraction The extraction fails
because no dye
normalization list is found
New Version - Yes False Run a feature extraction Feature Extraction uses
the dye normalization
probe selection method
set in the Correct Dye
Biases step of the
protocol.
Table 2 Dye normalization list rules
Grid template/design has dye normalization list association?
Protocol use dye norm list selection
Action Result
82 Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Setting Up and Changing Grid Templates and Files
Setting Up and Changing Grid Templates and Files
To show or hide the Grid Template Browser
Feature Extraction 12.2 User Guide
When you first open the Feature Extraction program after installation, the Grid Template Browser appears on the right- hand side of the main window.
If the Browser is closed, you can re- open it:
Click View > Grid Template Browser.
To add or change a grid template to an extraction set
If you add or change an array
design/grid template that does not
match the design file for an Agilent
image, the program highlights the
template in orange in the
Extraction Set Configuration tab to
warn you. Extraction can still be
run on this set, if you did not select
for Feature Extraction to
automatically assign a grid
template, but gridding or
annotation may be incorrect.
If you want to add an array design/grid template or use a different grid template with the image file in the extraction set, follow these instructions:
1 Select a grid template from the list of Grid Templates.
2 Drag and drop the grid template onto the extraction set.
The new grid template appears in the extraction set.
83
2 Extracting Microarrays Automatically To add or change a grid file to an extraction set
To add or change a grid file to an extraction set
84
Although you can add a grid file for Agilent microarrays, you have to add or create a grid file for non- Agilent microarrays.
1 In the Project Explorer, right- click the second line of the extraction set, and click Select grid file from the menu.
2 Select the .csv file that you want to use with the image file, and click Open.
To add or remove a grid template
To add an array design/grid template to the Grid Template Browser from a file:
1 Click Tools > Grid Template > Add.
Or
Right- click the Grid Template Browser, and click Add.
2 In the Add Grid Template dialog box, locate and select an array design file or grid file, and click Open.
Grid templates are created from either array design files or grid files. They do not exist as objects in the file system.
To remove a grid template/array design from the Grid Template Browser:
1 Click Tools > Grid Template > Remove.
Or
Right- click the Grid Template Browser, and click Remove.
2 Click Yes when asked if youre sure you want to remove the grid template from the browser.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To open grid mode from an extraction set
To open grid mode from an extraction set
Feature Extraction 12.2 User Guide
If automatic grid placement fails during extraction, you may want to create and use a grid file in place of the grid template, or edit the grid file, depending on which is associated with the extraction set. Or, you may just want to fine tune a grid placement. You can more easily access grid mode from Project Explorer in configuration mode than from the image display.
For non-Agilent images, you can
also access grid mode from the
image display. See Step 1. Start
grid mode from the non-Agilent
image file on page 136.
1 In Project Explorer, right- click the grid template or grid file in the extraction set.
2 Click Open Current Extraction in Grid Mode.
The image file for the extraction set appears in grid mode, ready for gridding.
See Chapter 3, Creating Grid Files in this guide to learn how to set up or edit a grid file.
To update array designs/grid templates online (via Agilent eArray)
NOTE For images produced on a NimbleGen, Molecular Devices, or Innopsys
scanner, the program first converts the image and then opens the image in
grid mode.
As new genomic information becomes available, Agilent updates its array design files (grid templates). These updates are available for downloading through the Agilent eArray website.
There are two different ways to load new array designs/grid templates that contain the latest annotation available on eArray into the Feature Extraction database.
Automatic updates of required and existing array designs/grid templates.
85
86
2 Extracting Microarrays Automatically To update array designs/grid templates online (via Agilent eArray)
You can set up Feature Extraction to automatically update array designs/grid templates that are required by the extraction sets, but are not available in the database. During the extraction run, the program logs into the eArray server using your eArray user name and password and download, install, and use the array designs/grid templates to process the extraction sets. See Setting up eArray Login for Automatic Updates.
Manual updates of new and existing array designs/grid templates.
You can manually update existing array designs/grid templates with the latest annotation or add new array designs/grid templates that are not in the database. Provide the user name and password of your eArray account.
To automatically update required and existing array designs/grid templates
1 Select Tools > eArray Settings.
The dialog box shown in Figure 15 appears.
Figure 15 eArray Login Setting window
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To update array designs/grid templates online (via Agilent eArray)
Feature Extraction 12.2 User Guide
2 Type your User Name and Password in the eArray Login Setting window, if you already have an eArray account.
Hint: The user name is the e- mail address you provided when you created an eArray account.
If you do not have an eArray account, register for one. Click Register to start the registration process. This takes you to the Agilent eArray registration website where you can request a free eArray account. Follow the instructions provided on the eArray website to complete the registration.
3 Click Test Connection. The message Login Successful!!! pops up if the connection to the eArray server is successful. If you get the message Login denied, check that the user name and password you typed are correct.
4 In the Advanced Options, select the appropriate boxes:
Use eArray server during extraction
Select this box to let Feature Extraction log into eArray server during the extraction run to download, install, and use required grid templates (array designs) that are missing in the Feature Extraction database.
Check for updates of grid template
Select this box if you want Feature Extraction to download updates to existing array designs/grid templates already in the database. Existing designs/grid templates in the database are not replaced, unless you select Replace old grid template.
Replace old grid template
Select this box to automatically replace old array designs/grid templates in the Grid Template Browser with the latest ones from eArray server.
87
88
2 Extracting Microarrays Automatically To update array designs/grid templates online (via Agilent eArray)
NOTE The existing grid templates/designs are not overwritten in the Feature
Extraction Database; the new version is added to the database without
deleting the previous version. However, Feature Extraction uses only the
latest version of a design/grid template, unless you designate the older
design/grid template version in the extraction set, and clear the Use
eArray server during extraction check box. See Setting up eArray Login
for Automatic Updates on page 36.
5 To save the eArray login settings, click Save.
You see a message Setting updated successfully. Click OK to close the message.
6 Click Close to exit the eArray Login Setting window.
To manually update new and existing array designs/grid templates
1 Right- click anywhere in the Grid Template Browser, and click Online Update.
Figure 16 appears with the design IDs and version dates of the array designs/grid templates that currently exist in the Feature Extraction database.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To update array designs/grid templates online (via Agilent eArray)
More than one version of the array
design/grid template can reside in
the database. The update function
lets you choose to add the latest
version from eArray to already
existing versions or to overwrite
the most recent version already in
the database.
Feature Extraction 12.2 User Guide
Figure 16 Grid Template Online Updating window
2 Click Choose All or select only the Design IDs that you want to check for updates, and the bottom buttons are activated.
3 If you want to add a Design ID not in the list, type the ID in the Add Design ID field, and click Add.
If you dont see your user name in the User name field, set up your eArray login first. See To automatically update required and existing array designs/grid templates on page 86 for details on how to set up your eArray login.
4 To check the release dates, click Check Latest Release Date, Or
5 To add the most recent array design/grid template to the database, click Update DB with New Release.
If you want to add the latest releases to the Feature Extraction database without overwriting the most recent version, clear the check box for Update Grid Template in FE database by replacing the most recent version.
89
90
2 Extracting Microarrays Automatically To display or change grid template properties
To display or change grid template properties
NOTE The existing grid templates/designs are not overwritten in the Feature
Extraction Database; the new version is added to the database without
deleting the previous version. However, Feature Extraction uses only the
latest version of a design/grid template, unless you clear the Use eArray
server during extraction check box in the eArray settings. See Setting up
eArray Login for Automatic Updates on page 36.
You can only change grid template
properties such as default
protocols and associated dye
normalization lists. Array design
properties (such as geometry)
cannot be changed.
1 Double- click the grid template/array design whose properties you want to display or change
2 Display the grid template information.
You can display all of the grid information. You can also change the default protocol (Agilent or non- Agilent 2- color, CGH, ChIP, or miRNA), the default 1- color protocol, and the dye normalization gene list.
Figure 17 Grid Template Properties window
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To add or change the default protocol in the grid template
To add or change the default protocol in the grid template
Feature Extraction 12.2 User Guide
1 Double- click the grid template whose protocol you want to add or change.
2 Click the cell to the right of Default protocol, and select the protocol from the list.
If you dont see the protocol you want from the list, import the protocol into the database first. See To import or export a protocol to or from the database on page 107.
To add or change the default one-color protocol in the grid template
Agilent design files and grid templates are the same for 2- color and 1- color or 2- color experiments. To make sure that the program can use a 1- color protocol with 1- color data, both 2- color and 1- color default protocols are specified. The program uses the 1- color protocol when the 1- color scan file is loaded.
1 Double- click the grid template whose protocol you want to add or change.
2 Click the cell to the right of Default OneColor protocol, and select a 1- color protocol from the list.
If you dont see the protocol you want from the list, import the protocol into the database first. See To import or export a protocol to or from the database on page 107.
91
2 Extracting Microarrays Automatically To change the dye normalization gene list in the grid template
To change the dye normalization gene list in the grid template
92
If the protocol specifies use of a dye normalization gene list and the extraction set includes a grid template, you create or select the gene list in the Grid Template Properties window.
1 Double- click the grid template whose dye normalization gene list you want to add or change.
2 Click the arrow to the right of the cell next to Dye Normalization Gene List.
Figure 18 Dye Normalization Gene List options
3 Select one of the following options:
Dye Normalization
Gene List
This is blank if no DyeNorm Gene Lists are available or if you select Set Empty.
If you previously created an internal dye normalization gene list, or selected an external gene list, select a name from this list. Once selected, you can double- click the gene list name to open the DyeNormList Editor to display or edit the list.
Create New
Browse file
Save As
Set Empty
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To create a new internal dye normalization gene list for a grid template
To create a new internal dye normalization gene list for a grid template
Feature Extraction 12.2 User Guide
1 Double- click the grid template for which you want to create a dye normalization gene list.
2 Click the cell to the right of Dye Normalization Gene List.
3 Click the arrow and click Create New.
NOTE You can create a dye normalization gene list using the Gene/Probe View,
or using the Chromosome View, but not both.
93
94
2 Extracting Microarrays Automatically To create a new internal dye normalization gene list for a grid template
Figure 19 DyeNorm List EditorGene/Probe View
Add genes to list by search criteria
To streamline the process of selecting genes for the list, you can search for genes according to Probe Name, Gene Name or Systematic Name. SNP probes are not included.
1 If necessary, click the Gene/Probe View tab.
2 Click the box next to Name and type a name for the dye normalization list.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To create a new internal dye normalization gene list for a grid template
Feature Extraction 12.2 User Guide
3 Under Criteria, click Search for, and then select Probe Name, Gene Name or Systematic Name as the type of name for the search.
4 Select how you want to use the name in the search. You can search according to the exact name, part of the name, or using a regular expression.
5 Type the name or part of the name for the search.
6 Select to include or exclude results.
7 Click Search.
Name
Features Selected (in %)
The dye normalization list has to contain at least 1% of the total features in the array. This indicates the % of the features currently selected for the list.
Grid Template Name
This shows the grid template with which you are associating this dye normalization list.
Contains Name (case sensitive)
If selected, you can search for part of the name, or you can use a regular expression. Match the case exactly. Regular expressions give you advanced search capabilities. For example,
If you want to search all probes you can type the regular expression as .(dot) to search all the probes.
Match Exact Name (case insensitive)
If selected, type the exact probe name, gene name or systematic name, (depending on the type of name you selected in step 1.) For this search, the case does not need to match.
Include Results
Exclude Results
95
96
2 Extracting Microarrays Automatically To create a new internal dye normalization gene list for a grid template
Add chromosomes to list
1 Click the Chromosome View tab.
Figure 20 DyeNorm List EditorChromosome View
2 Select the chromosomes from the Available list, and click Add>>. See Add/remove genes or chromosomes to list interactively on page 97.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To edit an internal dye normalization list
Feature Extraction 12.2 User Guide
Add/remove genes or chromosomes to list interactively
1 Select the genes or chromosomes of interest from the Available list.
For multiple selection, hold down the Shift key and click the first name and then the last name or hold the Ctrl key and click for non- contiguous names.
2 Click Add >>, or
Click <
If you attempt to transfer a large number of genes or chromosomes from list to list, the process can take time.
Save dye normalization gene list to internal file
You may have to refresh the screen
in order to see the name of the new
list in the Grid Template Properties
window.
1 Type a name for the dye normalization list in the Name field.
2 Click Save.
If you want to save the internal dye
normalization list as an external
file, see To export an internal dye
normalization gene list on
page 98.
When you click Save, a list is created using the Name specified, and is placed in the internal list available to the grid template.
To edit an internal dye normalization list
1 Double- click the grid template whose dye normalization gene list you want to change.
2 Click the cell to the right of Dye Normalization Gene List.
3 Select an existing gene list.
4 Double- click the list to bring up the DyeNormList Editor.
5 Add and remove any genes or chromosomes that you want.
97
2 Extracting Microarrays Automatically To import external dye normalization list to create a new internal list
Once an internal dye normalization
gene list is created, you cannot
change it. Save the changes using
a different name.
98
6 Click Save As.
7 Type a name for the edited list and click OK.
To import external dye normalization list to create a new internal list
1 Double- click the grid template whose dye normalization gene list you want to add or change.
2 If you want to create an internal list from an external .txt file, click the cell next to Dye Normalization Gene List, and click the down arrow.
3 Click Browse file, select the external .txt file and click OK.
This file requires a special format: 3 columns and a carriage return with Probe Name, Gene Name and Systematic Name listed in that order. A column headers row is not allowed.
4 Click the down arrow again, and select imported gene list as the default list.
If you need to make changes to the gene list, see To edit an internal dye normalization list on page 97.
To export an internal dye normalization gene list
After you create an internal dye normalization gene list file, you can export it as a .txt file to any folder you choose.
1 Double- click the grid template whose dye normalization gene list you want to export.
2 Click the down arrow to the right of the cell next to Dye Normalization Gene List, and select the internal gene list.
3 Click the down arrow again, and click Save As.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To create a new critical probes list for a grid template
Feature Extraction 12.2 User Guide
4 Select a folder you want to save the file to, rename the file, if necessary, and click Save.
The file is saved as an external .txt file.
To create a new critical probes list for a grid template
Feature Extraction 12.2 supports
the use of critical probes list. When
a critical probes list is associated
with a grid template, the program
notifies you when those probes are
flagged as outliers during the
extraction. A probe is only flagged
as an outlier if all the replicates of
that probe are outliers.
1 In the Grid Template Browser, double- click the grid template for which you want to create a critical probes list.
2 The FE GridTemplate Properties dialog box opens.
3 Click the cell to the right of Critical Probes List.
4 Click the arrow and click Create New.
The CriticalProbeList Editor dialog box opens.
99
100
2 Extracting Microarrays Automatically To create a new critical probes list for a grid template
Figure 21 CriticalProbeList EditorGene/Probe View
NOTE You can create a critical probes list using the Gene/Probe View, or using
the Chromosome View, but not both.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To create a new critical probes list for a grid template
Feature Extraction 12.2 User Guide
Add probes to critical probes list by search criteria
To streamline the process of selecting probes for the list, you can search for probes according to Probe Name, Gene Name or Systematic Name. SNP probes are not included.
1 If necessary, click the Gene/Probe View tab.
2 Click the box next to Name and type a name for the critical probes list.
3 Under Criteria, click Search for, and then select Probe Name, Gene Name or Systematic Name as the type of name for the search.
4 Select how you want to use the name in the search. You can search according to the exact name, part of the name, or using a regular expression.
5 Type the name or part of the name for the search.
6 Select to include or exclude results.
7 Click Search.
Name
Features Selected (in %)
This indicates the % of the features currently selected for the list.
Grid Template Name
This shows the grid template with which you are associating this critical probes list.
Contains Name (case sensitive)
If selected, you can search for part of the name, or you can use a regular expression. Match the case exactly. Regular expressions give you advanced search capabilities. For example,
If you want to search all probes you can type the regular expression as .(dot) to search all the probes.
Match Exact Name (case insensitive)
If selected, type the exact probe name, gene name or systematic name, (depending on the type of name you selected in step 1.) For this search, the case does not need to match.
Include Results
101
2 Extracting Microarrays Automatically To import a critical probes list
Exclude Results
102
If selected, all the results not found are transferred from the Available list to the Selected list.
Add/remove probes to list interactively
1 Select the probes of interest from the Available list.
For multiple selection, hold down the Shift key and click the first name and then the last name or hold the Ctrl key and click for non- contiguous names.
2 Click Add >>, or
Click <
If you attempt to transfer a large number of probes from list to list, the process can take time.
Save critical probes list to internal file
You may have to refresh the screen
in order to see the name of the new
list in the Grid Template Properties
window.
1 Type a name for the critical probes list in the Name field.
2 Click Save.
When you click Save, a list is created using the Name specified, and is placed in the internal list available to the grid template.
To import a critical probes list
1 In the Grid Template Browser, double- click the grid template to which you want to import a critical probes list from an external .txt file.
2 Click the cell next to Critical Probes List, and click the down arrow.
3 Click Browse file, select the external .txt file and click OK.
This file requires a special format: 3 columns and a carriage return with Probe Name, Gene Name and Systematic Name listed in that order. A column headers row is not allowed.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To export a critical probes list
Feature Extraction 12.2 User Guide
4 Click the down arrow again, and select the imported probe list as the default critical probes list.
To export a critical probes list
After you create a critical probes list file, you can export it as a .txt file to any folder you choose.
1 In the Grid Template Browser, double- click the grid template whose critical probes list you want to export.
2 Click the down arrow to the right of the cell next to Critical Probes List, and select the internal gene list.
3 Click the down arrow again, and click Save As.
4 Select a folder you want to save the file to, rename the file, if necessary, and click Save.
The file is saved as an external .txt file.
To edit a critical probes list
1 In the Grid Template Browser, double- click the grid template whose critical probes list you want to change.
2 Click the cell to the right of Critical Probes List.
3 Select an existing probes list.
4 Double- click the list to bring up the CriticalProbesList Editor.
5 Add and remove any probes that you want.
Once a critical probes list is
created, you cannot change it.
Save the changes using a different
name.
6 Click Save As.
7 Type a new name for the critical probes list and click OK.
103
2 Extracting Microarrays Automatically Working With Protocols
Working With Protocols
To move custom protocols from a previous version to version 12.2
104
In order to use custom protocols you created with a previous version of Feature Extraction (version 12.0 or 12.1) in Feature Extraction 12.2, first export those protocols from the previous version. Exported protocols are saved as .xml files. You can then import the saved protocol files to Feature Extraction 12.2. See To import or export a protocol to or from the database on page 107.
To show or hide the FE Protocol Browser
NOTE You can also use the Migration Utility to move grid templates, protocols,
and metric sets from Feature Extraction 12.0 or 12.1 to Feature Extraction
12.2. See the Migration Utility product documentation for instructions. See
the Feature Extraction 12.2 Release Notes for more information about the
Migration Utility and its availability.
When you first open the Feature Extraction program after installation, the FE Protocol Browser appears on the right- hand side of the main window.
If the Browser has been closed, you can re- open it:
Click View > FE Protocol Browser.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To add a protocol to an extraction set
To add a protocol to an extraction set
Feature Extraction 12.2 User Guide
By default, the program automatically determines the protocol to use for the extraction. If you want to manually add a protocol to an extraction set, follow these instructions:
1 Click to select a protocol in the FE Protocol Browser.
2 Hold down the mouse button and drag and drop the protocol onto the third line in the extraction set.
The new protocol appears in the extraction set.
To change the protocol used in an extraction set
If you want to use a different protocol than that associated with the grid template or project, follow these instructions:
1 Select a protocol from the FE Protocol Browser.
2 Hold down the mouse button and drag and drop the protocol onto the extraction set.
The new protocol appears in the extraction set, and replaces the previous protocol.
105
2 Extracting Microarrays Automatically To display or change protocol properties
To display or change protocol properties
106
1 Double- click a protocol in the FE Protocol Browser to open the Protocol Editor.
Figure 22 FE Protocol EditorProtocol Properties
2 Click the Protocol Properties button which shows the General tab.
Protocols and QC metric sets are
linked. Each protocol has a metric
set assigned to it. When Feature
Extraction updates protocols from
the eArray server, it also updates
the corresponding metric sets.
You can view information about the protocol. Under the General section, you can change the QC metric set that is associated with the protocol. Under the Steps section, you can remove protocol steps from the extraction run.
3 To change the QC metric set assigned to the protocol, click the cell to the right of Metricset, and select a QC Metric Set or Set empty.
4 To remove a protocol step from the run, clear the check box next to the step.
5 Click Save As to save the changes with a new protocol name.
See Chapter 4, Changing Protocol Settings for more details.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To import or export a protocol to or from the database
To import or export a protocol to or from the database
When importing a protocol from a
previous version of Feature
Extraction (i.e., version 12.0 or
earlier), the enhanced gridding
features are disabled by default.
For instructions on enabling the
enhancements, see To select to
use enhanced gridding on
page 196.
Feature Extraction 12.2 User Guide
To import a protocol to the database:
1 Click Tools > FE Protocol > Import.
Or
Right- click the FE Protocol Browser, and click Import.
2 Select a protocol from the list, and click Open.
To export a protocol from the database:
1 Click Tools > FE Protocol > Export.
OR
Right- click the FE Protocol Browser, and click Export.
2 Select a protocol from the list, and click Save.
A copy of the protocol is exported, but the protocol remains in the database.
To remove a protocol from the database
The Derived from message can
help you keep track of user-created
protocols. If the protocol is
Agilent-created, the message
indicates this. If the protocol is
user-created, the message lists the
original Agilent protocol from
which the new protocol was
derived.
1 Click Tools > FE Protocol > Remove.
Or
Right- click the protocol in the FE Protocol Browser, and click Remove.
2 Click Yes when asked to make sure that you want to remove this protocol.
Agilent- created protocols cannot be removed from the database.
107
2 Extracting Microarrays Automatically Working With QC Metric Sets
Working With QC Metric Sets
108
By default, the program calculates a set of default statistics that are displayed in the QC report. QC metric sets give you additional statistics for evaluating the Feature Extraction process on your microarrays. These statistics also appear in the QC report and in the Stats table of the text results file.
QC metric sets are linked with protocols. Each protocol has a metric set assigned to it. When the protocol is used by the program, the metric set is also used to generate metrics about the extraction set. During updates of protocols from the eArray server to the Feature Extraction database, the corresponding QC metric sets are also updated.
Agilent- created QC metric sets are installed with the Feature Extraction program and can be updated from the eArray website. You can also create your own QC metrics sets using the QC Chart Tool program. When you create a QC metric set in the QC Chart Tool program, it is added to the Feature Extraction Database.
If you wish to use your custom QC metric set with a protocol, associate it to the protocol. See To associate a QC metric set with a protocol on page 109.
To update the list of displayed QC Metric Sets
If the Feature Extraction program is open when you create a metric set, you need to refresh the QC Metric Set Browser in order to see the new metric set.
1 Click the right mouse button in the QC Metric Set Browser.
2 Select Refresh.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To move custom QC metric sets from a previous version to version 12.2
To move custom QC metric sets from a previous version to version 12.2
Feature Extraction 12.2 User Guide
If you created custom metric sets in a previous version of Feature Extraction (version 12.0 or 12.1), and you want to use them in Feature Extraction 12.2, export them from previous version. This saves each QC metric set in an .xml file. Then import the QC metric sets in Feature Extraction 12.2. See To import or export a QC Metric Set to or from the database on page 110.
To show or hide the QC Metric Set Browser
NOTE You can also use the Migration Utility to move grid templates, protocols,
and metric sets from Feature Extraction 12.0 or 12.1 to Feature Extraction
12.2. See the Migration Utility product documentation for instructions. See
the Feature Extraction 12.2 Release Notes for more information about the
Migration Utility and its availability.
This Browser contains the most recent Agilent QC metric sets after the program is installed. In addition, it shows custom QC metric sets that were created in the QC Chart Tool program. To show or hide the QC Metric Set Browser:
Click View > QC Metric Set Browser.
To associate a QC metric set with a protocol
You associate a QC metric set with a protocol in the FE Protocol Editor.
1 Double- click on the protocol to associate a QC metric set with. This opens the FE Protocol Editor.
2 Click the cell next to Metricset, and click the down arrow.
3 Select a QC metric set from the list.
109
2 Extracting Microarrays Automatically To import or export a QC Metric Set to or from the database
To import or export a QC Metric Set to or from the database
110
To import a QC Metric Set to the database:
1 Click Tools > QC Metric Set > Import.
Or
Right- click the QC Metric Set Browser, and click Import.
2 Select a QC Metric Set from the list, and click Open.
To export a QC Metric Set from the database:
1 Click Tools > QC Metric Set > Export.
Or
Right- click on a QC metric set in the QC Metric Set Browser, and click Export.
2 Select a QC metric set from the list, and click Save.
A copy of the QC metric set is exported, but the metric set remains in the database.
To remove a QC metric set from the database
Default Metric Sets and Protocols
that are loaded to the database
when you install the program are
not removable.
1 Click Tools > QC Metric Set > Remove.
Or
Right- click the QC Metric Set Browser, and click Remove.
2 Click Yes when asked to make sure that you want to remove this metric set.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To refresh the QC Metric Set Browser
To refresh the QC Metric Set Browser
Feature Extraction 12.2 User Guide
If you create a custom metric set using the QC Chart Tool program, it is added to the Feature Extraction database. If the Feature Extraction program was open when you created the metric set and you do not see the new metric set in your browser, use the following procedure.
1 In the QC Metric Set Browser, click the right mouse button
2 Select Refresh.
To view metric set properties
1 In the QC Metric Set Browser, double- click a metric set you wish to view.
For details on the logic used for
metric set evaluation, see the
Feature Extraction 12.2 Reference
Guide.
The QC Metric Set Properties for the selected metric set is displayed. Names of all the metrics included with this metric set are displayed, along with threshold values, if provided. For details on default metric set properties, see the Feature Extraction 12.2 Reference Guide.
111
112
2 Extracting Microarrays Automatically To view metric set properties
Figure 23 Metric set properties
2 Click Close to close the dialog box.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Running Feature Extraction
Running Feature Extraction
Feature Extraction 12.2 User Guide
Note the following requirements before you run Feature Extraction on a project.
If there are extraction sets in the project that contain a red X or (if the eArray connection is good) a lightning bolt, set up eArray login to let Feature Extraction automatically the download, install, and use grid templates, protocols, and QC metric sets that are missing from the database.
Internet connection is required during extraction, if Feature Extraction is expected to log into the eArray server to get the missing grid templates, protocols, and QC metric sets.
To start Feature Extraction
1 Click the Project Run mode On/Off icon in the toolbar, or click Project > Start Extracting.
If you did not save the project, a message appears that asks if you want to save the project. Save the project in order to run Feature Extraction.
2 Click Yes.
3 Click Save if you want to save the project to the default FE Project N.fep name, or type a new name, and click Save.
During extraction, information on what steps are performed is displayed in the Running Monitor at the bottom of the window. A Project Run Summary report is generated in the Summary Report tab when the extraction is complete.
To stop Feature Extraction
Click the Stop Extracting icon, or click Project > Stop.
113
2 Extracting Microarrays Automatically To create a new project with remaining extraction sets if run fails
To create a new project with remaining extraction sets if run fails
114
If a run fails due to a gridding problem, a message appears and provides an easy way to create a new extraction set containing the failed extraction sets. See Step 1. Create a new project for failed extraction sets on page 132 for details.
If a run fails for whatever reason, you can create a project containing the failed or not yet processed extraction sets left over from the previous project run. The following are requirements:
Project is a Standard FE project.
Project is in the Run mode.
Project just stopped extracting and contains extraction sets that failed or were not yet run.
1 Click Project > Create Project for Leftover Sets.
A message appears asking if you want to switch back to Config mode and continue.
2 Click Yes.
Feature Extraction creates a new project with all unfinished extraction sets from the current project. Feature Extraction also closes the project and starts the new project in its Config mode.
The name for the newly created project is Parent project name_2.
Save the new project before you extract or exit the project.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 Displaying Results
Displaying Results
An Agilent multiplex array is one
that has multiple Hyb/Wash
chambers for separate samples on
a single slide. These were formerly
referred to as multipack arrays.
Feature Extraction 12.2 User Guide
A Feature Extraction run on any image produces the following files:
A Project Run Summary report .rtf file
Text result .txt files (generated per microarray)
For Agilent arrays, a .pdf QC report (generated per microarray) or .html QC report (if you are running with a temporary license).
A .shp visual result file
NOTE For Agilent multiplex arrays, by default, text & QC reports are generated for
each subarray, but just one .shp and Project Run Summary is produced per
extraction of the image.
If you set the Crop multipack image to individual arrays option in the
project properties to True, the program separates the image into an
individual TIFF file for each subarray. The cropped images are saved in the
local folder. The original image is retained. Separate text and QC reports
are generated for each subarray.
This section describes how to view these results.
To display the text result file in Microsoft Excel
The result files are named with the following format:
Single pack: extraction set name_protocol name.
Multiplex (formerly known as Multipack): extraction set name_protocol name_pack number, indicated by row number.
1 Go to the folder specified for containing the results.
2 Right- click the text results file (protocol name on end), and click Open with.
3 Click Microsoft Excel.
115
116
2 Extracting Microarrays Automatically To display the text result file in Microsoft Excel
The text results open in the spreadsheet. The following figure shows all three tables (FEPARAMS, STATS, and FEATURE) in the file. You can also choose to create a different text file for each table. See To select to generate a single file for the text output on page 256 for more information.
Figure 24 Portion of text results file in Microsoft Excel
For some microarrays, the results do not all fit in an Excel spreadsheet (only the first 65536 rows in Excel 2003 and earlier). Results can be viewed in GeneSpring or DNA Analytics.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display the Project Run Summary Report
To display the Project Run Summary Report
Feature Extraction 12.2 User Guide
When you start the extraction run, the Summary Report tab appears showing the project start time. Once extraction is completed, you can see information about the extraction set in run summary. This summary is automatically saved as a Project Run Summary report that you can access later.
1 Go to the folder containing the project whose report you want to display.
The default directory for Feature
Extraction projects is C:\
ProgramData\Agilent\
FeatureExtractionStandalone\
FEProjects.
2 Double- click the file named YourComputerName_ LastBatchReport.rtf to display the last saved report for the project.
Or, double- click the file named FE Project N_timestamp.rtf to display a summary report for a specific run of the project.
Figure 25 Project Run Summary
117
118
2 Extracting Microarrays Automatically To display the Project Run Summary Report
You can open the QC Report by double- clicking the blue QC Report link in the Project Run Summary.
Evaluation of grid placement failures
If you see one of these warnings in the Project Summary Report, check the QC report and the .shp file to confirm the success or failure of the grid placement.
Grid failure error. There are many checks that are protocol dependent. These include number of features not found, number non- uniform local background regions, negative control standard deviation, and %CV of replicated features. Depending upon the thresholds set for these metrics in the protocol, the array may or may not show a grid failure error.
The average difference between the nominal grid location and the spot centroid is greater than 10 microns.
If the grid location does not fit within the image, the program stops the extraction and produces an error in the Project Run Summary.
If critical probes are found to be outliers
If you selected a critical probes list for the extraction, and one or more of the probes from the list is flagged as an outlier, then the summary report identifies those probes
If automated extraction fails for an Agilent image
Several options are available to troubleshoot the failure. See If Automatic Gridding for an Agilent Image Fails on page 132 for instructions for a successful grid placement and extraction.
If there is a warning that there is no barcode in the MAGE-ML output
Absence of a barcode may be due to several reasons:
Barcode reader failure
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display the QC Report
Feature Extraction 12.2 User Guide
Agilent scanner cannot read the Agilent microarray barcode for some reason.
Scan region set up to scan a non- Agilent microarray does not contain the barcode.
For Feature Extraction to load MAGE- ML results files into Rosetta Resolver, they must have a barcode or microarray identifier associated with them. You can add this identifier in the Scan Image Properties dialog box. See To add or change the New Array Identifier on page 263.
To display the QC Report
Six types of QC reports can be generated:
2- color GE QC report with or without spike- ins
1- color GE QC report with or without spike- ins
CGH_ChIP QC report
Streamlined CGH QC report
MicroRNA (miRNA) report
Chip test report (used by Agilent service personnel for scanner performance verification)
The report generated depends on the protocol used, and is specified in the Generate Results settings for the protocol. See Changing Protocol Step10: Generate Results on page 255.
1 Go to the folder specified for containing the results.
2 Double- click the QC Report .pdf or .html file.
The QC Report can be generated in
.pdf or .html format. If the Feature
Extraction license is a demo
license, only an .html file is
generated.
You can also access the QC Report from the Project Run Summary. Double- click the blue link next to QC Report. See Figure 25 on page 117.
119
120
2 Extracting Microarrays Automatically To display the QC Report
Figure 26 Streamlined CGH QC ReportPage 1
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To print the QC Report
To print the QC Report
Feature Extraction 12.2 User Guide
If the QC report is open in Internet Explorer, follow these instructions:
1 Click File > Page Setup.
2 Make sure that it prints in Portrait mode.
3 Select A4 size paper.
4 Click Printer > Properties, and select for color printing.
5 Exit until you see Page Setup, then close Page Setup.
6 Click File > Print Preview.
7 Adjust the margins to display all the images.
8 Click Print.
You can alternatively use another Internet browser that has a Shrink to Fit option for printing, such as Mozilla Firefox.
To display visual results
After extraction is completed, Feature Extraction generates one visual shape result file (.shp) for each tiff image file (.tif). For multiplex Agilent microarrays (more than one microarray on a single glass slide), a single .shp file is for the entire tiff image and covers all the arrays on the slide.
The shape file is only viewable in Feature Extraction. It provides visual results about the extracted microarray image. You can view extraction results such as feature cookies, gene names, raw signals, log ratios, outliers, and pixels used to quantify features (available for high- resolution images).
Load the visual results file (.shp) from Project Explorer
1 Right- click a processed extraction set.
2 Click View Visual Result.
The image is loaded, along with its shape file into the Image workspace. If no shape file is present, the menu item is grayed out.
121
122
2 Extracting Microarrays Automatically To display visual results
Load the visual results file (.shp) from the Image workspace
For more details on how to load the
shape file from the Image
workspace, see Loading Feature
Extraction Visual Results on
page 264.
1 Click Open Image, select the .tif file, and click Open.
You can also open the image from Project Explorer (right- click the image name and select View Visual Result) or the Extraction Set Configuration tab sheet (double- click the Scan File Name).
2 Click Feature Extraction > Load Visual Result.
3 Select the .shp file associated with the image, and click Open.
Work with the visual results
By default, the when you load a visual results file, Feature Extraction shows the image in View Outliers Only mode, displaying information about outlier features only. To view results for non- outlier features, see View results on low- resolution images: on page 123, and View results on high- resolution images (2- by 3- micron or 20- bit): on page 123.
Figure 27 View Extraction ResultsDefault Settings
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display visual results
Feature Extraction 12.2 User Guide
View results on low-resolution images:
1 After .shp file is loaded to the image, zoom in until you see enough of the spots to help you evaluate the results.
2 Click the cropping mode icon to turn off crop mode.
The cursor turns from the cropping cursor to the arrow cursor.
3 Click View > Extraction Results, and clear the View Outlier Only check box.
4 On the image, pass cursor to feature of interest and hold for a few seconds.
A description of the feature appears.
To find a specific feature, use Ctrl-F.
Type the Feature number of the
specific feature you want to find.
Figure 28 Visual results for extracted microarray image
View results on high-resolution images (2- by 3- micron or 20-bit):
After you first load the image, you, in fact, are viewing a preview image. You can only view the original high- fidelity image if you crop the size of the image to one- tenth or less of the original size.
1 After .shp file is loaded to the image, zoom in until you see enough of the spots to help you identify the region of interest.
2 Click the left mouse button at the point where you want to start forming a rectangle for cropping.
123
124
2 Extracting Microarrays Automatically To display visual results
3 With the crop cursor, draw a rectangle equal to or less than one- tenth the size of the image.
Note that the crop cursor is black if the size is within one- tenth (200% to 1000%) of the original size, but it turns red if you are drawing outside the high fidelity range.
4 Then release the mouse button.
Note the status bar at the bottom of the window. It now reads High Fidelity.
5 Click View > Extraction Results, and clear the check box View Outlier Only.
6 On the image, pass cursor to feature of interest and hold for a few seconds.
A description of the feature appears like in Figure 28 on page 123.
View Pixels Used
This option is available for low- resolution and high- resolution images that are in High Fidelity mode (i.e. when the image is zoomed in to one- tenth of its original size).
The View Pixels Used option displays the pixels used to quantify the feature signal within the cookie area. Pixels that are determined to be outliers are shown in black.
1 Complete steps 1- 5 from View results on high- resolution images (2- by 3- micron or 20- bit): on page 123.
2 Under View > Extraction Results, select View Pixels Used.
Feature Extraction 12.2 User Guide
Extracting Microarrays Automatically 2 To display visual results
Feature Extraction 12.2 User Guide
Figure 29 View Pixels Used results on a high-resolution image in
high-fidelity mode
125
126
2 Extracting Microarrays Automatically To display visual results
Feature Extraction 12.2 User Guide
Agilent Feature Extraction 12.2 User Guide
3 Creating Grid Files
About Grid Templates and Grid Files 128
When to Use Manual Gridding 129
Analyzing the Initial Grid for an Agilent Image 130
If Automatic Gridding for an Agilent Image Fails 132
Creating and Using a Grid File for non-Agilent Images Overview 136
Starting Grid Mode for a non-Agilent Image 139
Viewing and Changing Grid Geometry 149
Optimizing Grid Placement Using the Help Wizard 151
Manual Grid Adjustments 160
Saving a Grid File 165
Calculating Spot Size and Spot Centroids 166
Locating Spot Centroids Manually 167
Refitting a Grid 174
Changing Default Settings for Grid Editing 176
Shortcuts to Help You Work with Grids and Find Spots 177
Agilent Feature Extraction program uses grid templates and grid files to determine where spots are located on a microarray before extraction.
This chapter shows you how to troubleshoot failed Agilent image extractions and create grid files for Agilent and non- Agilent images.
Each section of this chapter describes the actions you take to properly position a main grid, subgrids, and ultimately the spot centroids. The section also gives you background information to help you position these grids and spot centroids.
127Agilent Technologies
3 Creating Grid Files About Grid Templates and Grid Files
About Grid Templates and Grid Files
A grid template contains spacing
information for the specified format
as well as a unique set of probes
and annotations for the microarray
design that it matches.
128
Feature Extraction automates the process of locating microarray spots through the use of grid templates or grid files, which contain spacing information required to place grids for specific Agilent microarray designs. Grid templates are derived from array design files of the same name that reside in the shared Feature Extraction database and are displayed in the Feature Extraction Grid Template Browser. When you import, add, or update a grid template, you actually import, add, or update the entire array design file.
A grid file is similar to a grid
template, except it contains
information specific for a single
scan of a given microarray slide.
If gridding fails for a particular Agilent microarray, create an Agilent grid file to use for the extraction. Agilent grid files are saved using the image name followed by _grid.csv. If you specify a grid file in a project for extraction of an Agilent image, also specify the grid template to use for the rest of the extraction steps.
For non- Agilent microarrays in a project, create an Agilent grid file for the specific non- Agilent microarray you intend to extract. A non- Agilent grid file is stored as a .csv (comma- separated value) file and is not saved in the database. When you save a non- Agilent grid file, two .csv files are created that together define the grid. They are named using the image name followed by _grid.csv or _feat.csv. Both files are required to define the grid.
You cannot, nor do you need to,
adjust spot centroids for Agilent
images because Agilent array
printing is accurate once an initial
grid is found.
When you create a non- Agilent grid file, locate the spot centroids, or geometric centers, on the microarrays. To locate spot centroids, you interactively position a main grid and subgrids on the microarray. After the subgrids are well- placed, you calculate spot size and centroids.
The program algorithm locates most, if not all, of the spots where you want them. Any that are incorrect can be moved using a Spot navigator (see Locating Spot Centroids Manually on page 167). This capability ensures that the Feature Extraction program identifies the correct feature and local background.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 When to Use Manual Gridding
When to Use Manual Gridding
Feature Extraction 12.2 User Guide
Table 3 Uses of manual gridding
Image type Reason to use manual gridding See
Agilent Image Look at how the grid file or template has
placed the grid
Analyzing the Initial Grid
for an Agilent Image on
page 130
Agilent Image If a gridding error appears in the Summary
Report after Feature Extraction
If Automatic Gridding for
an Agilent Image
Fails on page 132
Non-Agilent Image Create a grid file to use with the image for
Feature Extraction
Creating and Using a
Grid File for non-Agilent
Images Overview on
page 136
Agilent or non-Agilent image Grid is marked as unplaced in Extraction
Set
Optimizing Grid
Placement Using the Help
Wizard on page 151
Non-Agilent Image Evaluation of grid placement Starting Grid Mode for a
non-Agilent Image on
page 139
non-Agilent image Spot spacing or location of centroids is not
correct
Calculating Spot Size
and Spot Centroids on
page 166
Locating Spot Centroids
Manually on page 167
Agilent or non-Agilent image Grid placement is not correct after automatic
placement
Manual Grid
Adjustments on
page 160
Agilent (rarely) or non-Agilent image Grid does not fit the spots on the grid
(spacing is incorrect in the Geometry)
Refitting a Grid on
page 174
129
3 Creating Grid Files Analyzing the Initial Grid for an Agilent Image
Analyzing the Initial Grid for an Agilent Image
130
When you add an Agilent image to a project, the program assigns a grid template, based on the barcode (AMADID) and a protocol. The protocol uses the grid template to find the spots automatically. Normally, you do not need to do any grid adjustments for an Agilent image. However if you want to display the placement of the grid, or if the grid is labeled as unplaced you can open the extraction in grid mode and adjust it.
To open the current extraction in grid mode
For Agilent images, the grid
template associated with the
image is automatically assigned to
the extraction set. A grid is
assigned when the grid template is
available in the Grid Template
Browser, or if you configured
eArray to download templates
automatically.
1 Open a project file, or add an extraction to an existing project file.
2 If no grid is open, drag and drop a grid template from the Grid Template Browser. Or, right click on the grid icon and click Select Grid File to select an existing grid file.
3 Select a protocol.
Select a protocol that matches the image for grid mode. If not, the grid and spot placement may not be correct.
4 In the Project Explorer, right- click on the extraction set and select Open Current Extraction in Grid Mode.
You can toggle back and forth from
grid mode to the project explorer
from the Window menu on the
toolbar.
The program attempts to place the grid and the image opens in grid mode. In grid mode, you can view the grid placement and adjust if necessary.
See Optimizing Grid Placement Using the Help Wizard on page 151.
NOTE For images produced on a NimbleGen, Molecular Devices, or Innopsys
scanner, the program first converts the image and then opens the image in
grid mode.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To open the current extraction in grid mode
Feature Extraction 12.2 User Guide
Figure 30 Extraction opened in grid mode
Main Grid View
Main Grid Corner ViewGrid Adjustment pane Help Wizard
131
3 Creating Grid Files If Automatic Gridding for an Agilent Image Fails
If Automatic Gridding for an Agilent Image Fails
132
If the gridding fails during feature extraction, you see a red error message in the Summary Report that indicates why the gridding failed. See To display the Project Run Summary Report on page 117 for details on summary reports. If gridding fails, follow the following steps to adjust the grid and rerun an extraction set.
Step 1. Create a new project for failed extraction sets
Step 2. Access grid mode from a project extraction set
Step 3. Position grids and save grid file
Step 4. Set up and rerun Feature Extraction
Step 5. Adjusting rotation and skew of subgrids
Step 1. Create a new project for failed extraction sets
When you attempt to close a project or exit the Feature Extraction program after an extraction where one or more extractions failed due to a gridding error, the program asks if you want to create a new project containing the failed extraction sets.
Figure 31 Failed extractions due to gridding error
If you click No, the project or Feature Extraction program closes.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 1b. Manually create a new project for failed extraction sets
Feature Extraction 12.2 User Guide
If you click Yes, the program
Creates a grid file for all the failed extractions (due to gridding error)
Creates a project for all failed extractions (due to gridding error)
Assigns created grids to their respective extraction sets automatically, and their status is changed to unplaced.
The name for the newly created project is Parent project name_2. Every time a new project from leftover sets is created, the program appends _2 to the parent name.
Save the new project before you extract or exit the project.
An unplaced grid is one whose x-
and y- coordinates, skew, and
rotation are equal to zero.
When you run the project containing the failed extraction sets, Feature Extraction checks the grid placement status of the first extraction set. If the grid is unplaced a message appears asking you if you want to open grid mode. Click Yes to open the extraction in grid mode where you can place and save the grid. If you click No, the program checks the next extraction set in the projectuntil all extraction sets have been checked.
Step 1b. Manually create a new project for failed extraction sets
If a run fails for any reason, you can create a project containing the failed extraction sets left over from the previous project run. The following are required:
Project is a Standard FE project
Project is in the Run mode
Project has stopped extracting and contains failed extraction sets or extraction sets not yet run.
1 Select Project > Create Project for Leftover Sets.
A message appears asking if you want to switch back to Config mode and continue.
2 Click Yes.
133
134
3 Creating Grid Files Step 2. Access grid mode from a project extraction set
Feature Extraction creates a new project with all unfinished extraction sets from the current project. Feature Extraction also closes the current project and starts the new project in Config mode.
The name for the newly created project is Parent project name_2. Every time a new project from leftover sets is created, the program appends _2 to the parent name.
Save the new project before you extract or exit the project. To proceed, open the extraction set in grid mode by following Step 2.
Step 2. Access grid mode from a project extraction set
This method is the quickest way to access grid mode after automatic grid placement has failed. It is the only way to access grid mode for Agilent images.
After gridding, the extraction set
consists of the image file, the grid
template and grid file, and the
protocol.
In the Project Explorer on the left side of the project window, right- click on any object in the extraction set (extraction set name, image, grid template or file, protocol).
3 Select Open Current Extraction in Grid Mode.
The image file for the extraction set appears in grid mode, ready for positioning of the main grid.
Also, the grid file appears under the original grid template in the extraction set.
Step 3. Position grids and save grid file
1 Open the extraction in grid mode.
2 Follow the help wizard steps to:
a Position the main grid.
The spot size and spot centroid
options in the Help Wizard are
used for non-Agilent images only.
b Position subgrids, if necessary.
See Optimizing Grid Placement Using the Help Wizard on page 151.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 4. Set up and rerun Feature Extraction
The grid is labeled as unplaced in
the Project Explorer until the grid is
positioned and grid mode is closed.
Feature Extraction 12.2 User Guide
3 Save the grid file.
See Saving a Grid File on page 165.
Step 4. Set up and rerun Feature Extraction
1 Close the grid mode window.
2 Save the project.
3 Click the Project Run mode On/Off icon to start the extraction.
135
3 Creating Grid Files Creating and Using a Grid File for non-Agilent Images Overview
Creating and Using a Grid File for non-Agilent Images Overview
136
Follow these instructions to create and use a grid file for non- Agilent images. A grid file is required for non- Agilent images to run Feature Extraction.
Step 1. Start grid mode from the non- Agilent image file.
Step 2. Position grids, locate spots and save grid file.
Step 3. Extract the image file with its new grid file.
Step 1. Start grid mode from the non-Agilent image file
1 Open the non- Agilent image file.
See To open an image file on page 139.
Grid mode cannot be accessed
from an open Agilent image. For
Agilent images, access grid mode
from within an extraction set. See
Step 2. Access grid mode from a
project extraction set on
page 134.
2 Open grid mode and select the gene list type and protocol to use.
See To open grid mode for a non- Agilent image on page 142.
3 Click OK.
4 In the Geometry Information dialog box, click Auto Fit.
The image file for the extraction set appears in grid mode, ready for positioning of the main window. See Geometry information on page 148.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 2. Position grids, locate spots and save grid file
Step 2. Position grids, locate spots and save grid file
Feature Extraction 12.2 User Guide
1 Follow the help wizard steps in the lower left corner of the grid mode window to:
If you are using a non-Agilent
microarray, you decide to adjust
spots, calculate spot sizes and
centroids, or position spots
manually. You base this decision on
visual inspection of how the grid is
fitted.
a Determine nominal spot and subgrid spacing manually. (Optional and only for third- party arrays.)
b Determine nominal subgrid spacing. (Optional and for third- party arrays with subgrids only.)
c Position main grid and subgrids, if necessary.
See Optimizing Grid Placement Using the Help Wizard on page 151.
The spot size and spot centroid
options are used for non-Agilent
images only.
2 Click the Adjust Spots icon to enable spot location adjustment.
3 Click the Calculate spot size and centroids icon and calculate the spot sizes and centroids.
See Calculating Spot Size and Spot Centroids on page 166.
4 Visually inspect the spot centroids and if necessary, position spots manually.
See Locating Spot Centroids Manually on page 167.
5 Save the grid file.
See Saving a Grid File on page 165.
137
3 Creating Grid Files Step 3. Extract the image file with its new grid file
Step 3. Extract the image file with its new grid file
138
To test the new grid file, create the project and add the extraction set.
1 Select Feature Extraction > Create a New FE Project.
A new Feature Extraction project is created with the image file already added.
2 In the Project Explorer, right- click on the grid icon below the image file, and click Select Grid File.
3 In the Select Grid file for Extraction Set dialog box, select your saved grid file, and then click Open.
4 From the FE Protocol Browser, drag the protocol for the image type to the extraction set.
5 Save the project and run Feature Extraction to see if the grid works.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Starting Grid Mode for a non-Agilent Image
Starting Grid Mode for a non-Agilent Image
Feature Extraction 12.2 User Guide
You can view and adjust grid placement using grid mode. For non- Agilent images, grid mode can be accessed from either an open image file or from an extraction set that has an associated grid and protocol.
To open an image file
You cannot access grid mode for
Agilent images from an open
image. See Step 2. Access grid
mode from a project extraction
set on page 134.
1 Click on the toolbar Open image file button, or click File > Open > Image, to display the Open Dialog Box.
2 In the Open Dialog Box, double- click a .tif file from the scanner, or select a file and click Open.
You can also open an image file by
dragging the file to the Feature
Extraction desktop icon.
An image of the scanned microarray contained in the file opens in its own window.
Note that for images produced on a NimbleGen, Molecular Devices, or Innopsys scanner, the program first converts the image before opening it.
139
140
3 Creating Grid Files To open an image file
Figure 32 TIFF image for non-Agilent microarray
Toggle Log Scale Icon
Grid Mode IconZoom In/Zoom Out Icons
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To open an image file
Feature Extraction 12.2 User Guide
3 To see the spots more clearly, click the Toggle Log Scale icon and click the Zoom In icon to increase the image size.
Figure 33 Image view with log scale on and image zoomed
141
3 Creating Grid Files To change the Portrait/Landscape view of the image
To change the Portrait/Landscape view of the image
The program changes orientation
in this way because the Agilent
scanner scans the long side of the
microarray through the glass. This
orientation is flipped and rotated
compared with other scanners.
142
If the gene list file with the grid information organizes its data with the opposite orientation, change the orientation of any opened image. You typically do not have to change orientation for Agilent images, but you usually have to for non- Agilent images.
1 Select Tools > Flip Upper Left to Lower Right (Landscape/Portrait).
The program takes the upper left corner of the image and flips it to the lower right. It does not matter whether or not you change the orientation.
2 Select File > Save modified image to save the image.
To open grid mode for a non-Agilent image
Images produced on a NimbleGen, Molecular Devices, or Innopsys
in grid mode.
1 Open a non- Agilent image file.
2 Click the Grid button on the image toolbar.
The Create/Edit Grid dialog box opens.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To open grid mode for a non-Agilent image
Feature Extraction 12.2 User Guide
Figure 34 Create/Edit Grid dialog box
3 Select the Gene List Type to let you position the grid.
If you select a file whose data orientation is different from the image, the Agilent cannot set the grid properly. Cancel this dialog box and change the orientation if you have to.
143
3 Creating Grid Files To open grid mode for a non-Agilent image
Agilent Grid File (*_grid.csv)
144
A grid file that has been saved from the previous creation of a grid on an image.
If you want to edit a grid or change the location of the spot centroids for an image, select the grid file previously saved for that image.
If you have never set up an initial grid and the image resembles one whose grid you have set up before, select that grid file from the already existing grid files.
Gene List
The requirements for layout information and feature annotations include a tab- delimited text format that contains at least these columns: Grid index (or SubGridRow/SubGridCol), SpotRow, SpotCol and GeneName or Probename.
The tab text file can be a GAL file, a results file, or an Excel file exported as a tab- delimited text file.
Control Type can also be in the tab text file. See the Feature Extraction 12.2 Reference Guide to see the definitions of the control types that can appear in the tab text file.
No Gene List
4 Select a protocol.
For 2 color Gene Expression, the
default
GE2-NonAT_
protocol provided with the program
can be used for non-Agilent
images.
Select a protocol to help with grid placement and spot finding. Grid placement and spot finding algorithms differ between protocols. Select a protocol that matches the image for grid mode. If not, the grid and spot placement will probably not be correct.
5 Click OK.
What happens next depends on what you have selected for the gene list type:
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To open grid mode for a non-Agilent image
Feature Extraction 12.2 User Guide
If you select Agilent Grid File as the gene list type
A Browser appears that lets you select the grid file. Make sure that the grid file already exists for the image.
Select a grid file, and click Open.
The Main Grid View appears with the grid approximately positioned.
If you select Gene list and then select a gal file as the gene list type
The Geometry Information dialog box appears for the gal file grid.
1 Change the default values if necessary.
2 Click Auto Fit or Manual Fit.
For a description of all the fields and buttons in the Geometry Information dialog box, see Geometry information on page 148.
The Main Grid View for this image appears.
If you select Gene list and then select a tab text file as the gene list type
The Tab Text Grid File Preview window appears.
145
146
3 Creating Grid Files To open grid mode for a non-Agilent image
Figure 35 Tab Text Grid File Preview window with header selected
1 Click each of the column headers that correspond to the following information fields:
Subgrid Col
Subgrid Row
Spot Col
Spot Row
Probe Name
2 Select the appropriate name from the list.
Change the names of the columns listed in order for gridding to work with a tab text file. You can also change other column names, such as Gene Name or Control.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To open grid mode for a non-Agilent image
Feature Extraction 12.2 User Guide
Figure 36 Tab Text Grid File Preview window with headers selected
3 Click OK.
The Geometry Information dialog box appears.
4 Click Auto Fit or Manual Fit.
For a description of all the fields and buttons in the Geometry Information dialog box, see Viewing and Changing Grid Geometry on page 149.
The Main Grid View for the non- Agilent image appears.
147
148
3 Creating Grid Files To open grid mode for a non-Agilent image
If you select No Gene List
The Geometry Information dialog box appears.
1 Change the default values if necessary.
2 Click Auto Fit or Manual Fit.
For a description of all the fields and buttons in the Geometry Information dialog box, see Viewing and Changing Grid Geometry on page 149.
The Main Grid View for the non- Agilent image appears.
Geometry information
This dialog box appears when you click OK from the Create/Edit Grid dialog box.
If Gene List or No Gene List are selected as the Gene List Type in the Create/Edit Grid dialog box, the Feature Extraction program attempts to discover the grid and populates the Geometry Information fields with the information from that attempt.
For Gene List, if you select a .gal file, the parameters from the .gal file is displayed in the Geometry Information dialog box.
For details on all of the fields and buttons of the Geometry Information dialog box, see Viewing and Changing Grid Geometry on page 149.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Viewing and Changing Grid Geometry
Viewing and Changing Grid Geometry
Feature Extraction 12.2 User Guide
The following dialog box appears when you click OK from the Create/Edit Grid dialog box when entering grid mode from a non- Agilent image file (see Starting Grid Mode for a non- Agilent Image on page 139). It also appears when you Refit a grid (see Refitting a Grid on page 174).
Figure 37 Geometry Information dialog box
Main grid/Array
geometry
For the main grid, enter dimensions in the form of number of columns and rows. Or, enter the distance between subgrids in pixels for both columns and rows.
Subgrid geometry
For subgrids, enter dimensions in the form of number of columns and rows, or enter the distance between spots in pixels for both columns and rows.
149
3 Creating Grid Files Viewing and Changing Grid Geometry
Double Density Offsets
150
The double density offset is applied when the array being gridded contains features consisting of two interlaced rectangular grids. The spacing of these rectangular grids is specified in the Double Density row and column offsets.
Nominal spot diameter
Because spots may not be circular, enter the horizontal and vertical dimensions in pixels.
Reset
Auto Fit
Manual Fit
Cancel
Table 4 Geometry information that you can change on the Geometry Information (GI) dialog box*
* You can change the nominal spot and subgrid spacing in grid mode.
File Type Initial GI dialog box displays?
Change main grid dimensions?
Change subgrid dimensions?
Change spot spacing?
Change spot size?
Agilent Grid No
The Geometry Information dialog box does not appear when you select Agilent Grid File as the gene list type. The columns for the Agilent Grid file type apply to the refit Geometry Information dialog box, which does appear when you attempt to refit an Agilent grid (see Refitting a Grid on page 174).
No No Yes No
GAL Format Yes Yes No Yes Yes
Tab Text Yes Yes No Yes Yes
No gene list Yes Yes Yes Yes Yes
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Optimizing Grid Placement Using the Help Wizard
Optimizing Grid Placement Using the Help Wizard
Although the Help Wizard is
available for both Agilent and
non-Agilent arrays, some steps
apply only to non-Agilent
(third-party) arrays.
Feature Extraction 12.2 User Guide
Grid mode provides a Help Wizard to step you through the process of optimizing the placement of the grids. To view the help steps, in the Help pane on the lower left of the grid mode window, click Next. To go back to the previous step, click Previous. The help wizard provides instructions for the following tasks:
Step 1. Get nominal spot spacing manually (for non- Agilent arrays only)
Step 2. Get nominal subgrid spacing manually (for non- Agilent arrays only)
Step 3. Adjust the position of the main grid
Step 4. Adjust the subgrid
Step 5. Adjusting rotation and skew of subgrids
Step 6. Save grid file
To zoom in or out of the views
Before you optimize the grid placement, zoom the image in order to see the spots more easily. There are two ways to zoom the image:
See To change the magnification
(zooming) on page 279 of this
guide for more details on zooming.
From the Feature Extraction menu bar, select View > Zoom and choose a % zoom from 10% to 1000%. A 10% zoom is defined and you can select a zoom level in 10% increments. You can also change the default zoom settings. See Changing Default Settings for Grid Editing on page 176 for details.
On the image toolbar, click one of the zoom icons.
151
3 Creating Grid Files To undo or redo grid position movements
Figure 38 Grid mode view
To undo or redo grid position movements
Main Grid View Click to zoom out.
Corner View
Click to zoom to 100%.
Grid Adjustment pane Help Wizard
Click to zoom in.
Image Toolbar
152
Click the Undo button, , to undo an action.
Click the Redo button, , to redo an action.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 1. Get nominal spot spacing manually
Step 1. Get nominal spot spacing manually
Feature Extraction 12.2 User Guide
This step is optional, and is only used for third- party arrays. Perform this step if spacing information for spots is unknown.
1 Make sure that the color log scale has been turned on to view the corner spots on the main grid more easily.
2 On the gridding toolbar, click the View and Measure Image icon.
3 If necessary, zoom the image. See To zoom in or out of the views on page 151.
4 On this image view (which is appended with [CROPPED] in the image name), click on the center of the top left spot and drag the mouse horizontally to the center of the spot in the next column. (Do not release the mouse button.) Read and note the column distance from the status bar. This number is the left- hand number shown in parentheses; for example, (16x0). (You need this value later.)
5 On the image view, click the center of the top left spot and drag the mouse vertically to the spot center in the first column of the second row. (Do not release the mouse button.) Read and note the row distance from the status bar. This number is the right- hand number shown in parentheses; for example, (0x17). (You need this value later.)
6 Close the cropped image view.
7 In the grid mode view, click the Grid Definition tab.
8 In the Grid Definition panel, double- click the space next to Col under Nominal Spot spacing, and enter the column distance value from step 4.
9 In the Grid Definition panel, double- click the space next to Row under Nominal Spot spacing, and enter the row distance value from step 5.
153
154
3 Creating Grid Files Step 1. Get nominal spot spacing manually
Figure 39 Reading column or row distance from image view
Read column (horizontal) distance
Read row (vertical) distance
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 2. Get nominal subgrid spacing manually
Step 2. Get nominal subgrid spacing manually
Feature Extraction 12.2 User Guide
This step is optional and only for third- party arrays. Perform this step if spacing for the subgrids is unknown.
1 Make sure that the color log scale has been turned on to view the corner spots on the main grid more easily.
2 On the gridding toolbar, click the View and Measure Image icon.
3 If necessary, zoom the image. See To zoom in or out of the views on page 151.
4 On the newly opened image, click the center of the top right spot for the first subgrid and drag the mouse horizontally to the center of the first spot in the adjacent subgrid. (Do not release the mouse button.) Read and note the column distance from the status bar. This number is the left- hand number shown in parentheses; for example, (43x0). (You need this value later.)
5 On the opened image view, click the center of the bottom left spot of the first subgrid and drag the mouse vertically to the center of the top left spot of the subgrid under it. (Do not release the mouse button.) Read and note the row distance from the status bar. This number is the right- hand number shown in parentheses; for example, (0x42). (You need this value later.)
6 Close the opened image view.
7 In the grid mode view, click the Grid Definition tab.
8 In the Grid Definition panel, double- click the space next to Col under Nominal Subgrid spacing, and enter the column value from step 4.
9 In the Grid Definition panel, double- click the space next to Row under Nominal Subgrid spacing, and enter the row value from step 5.
155
3 Creating Grid Files Step 3. Adjust the position of the main grid
Step 3. Adjust the position of the main grid
156
1 Make sure that the color log scale has been turned on to view the corner spots on the main grid more easily.
2 On the toolbar, click the Adjust Main Grid icon.
3 Move the mouse into the upper Main Grid View until the move cursor appears.
4 Click and drag the move cursor so that the top left corner of the grid is positioned on the top left spot of the microarray image.
5 To do fine adjustment of the main grid, click the top left spot in the top left pane of the lower Corner View pane.
The thick line represents the main
grid and the thin line represents
the subgrids.
Figure 40 Adjust main grid view
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 4. Adjust the subgrid
Step 4. Adjust the subgrid
For a microarray with only one grid,
if the main grid is well-positioned,
there is no reason to move the
subgrid. This is the case for most
Agilent microarrays.
Feature Extraction 12.2 User Guide
When you position subgrids you are, in effect, positioning nominal spot centroids (+). These nominal spot centroids are still subject to adjustment by the spot finding algorithm in Feature Extraction. You can position the subgrids such that the nominal spot centroids are just inside the spots. These positions are good enough for the spot finding algorithm to calculate more accurate spot centroids.
1 On the toolbar, click the Adjust Subgrid icon.
2 In the Main Grid View, click on a subgrid.
3 In the Main Grid View, click and drag the top left corner of the subgrid to the top left spot of the subgrid.
4 For fine adjustment of the grid, click the Select Corners and Automatically Fit Grid icon on the toolbar to turn it off.
5 In lower Corner View, click the center of the top left spot of the selected subgrid.
6 Repeat step 2 through step 5 for the remaining subgrids.
157
158
3 Creating Grid Files Step 4. Adjust the subgrid
Figure 41 Adjust subgrid view
Subgrid Main Image View Subgrid Corner View
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Step 5. Adjusting rotation and skew of subgrids
Step 5. Adjusting rotation and skew of subgrids
Feature Extraction 12.2 User Guide
1 In the Main Grid View, click on a subgrid to select it.
When Select Corners and Automatically Fit Grid is enabled,
the program automatically
calculates a fit based on three
user-selected corner points.
2 On the toolbar, click to select the Select Corners and Automatically Fit Grid icon, if it is not on.
3 In the lower view, if the corners of the grid are not near to the corner spots of the image, adjust the grid location by clicking and dragging the grid in the bottom right pane of the lower four corner view.
4 Move the cursor to the top left pane of the lower Corner View. Notice that the cursor turns to a red cross. Click on the top left spot of the subgrid.
5 In the top right pane of the lower Corner View, click on the top right spot (in the first row) of the subgrid.
6 In the bottom left pane of the lower Corner View, click on the lower left spot (in the first column) of the subgrid.
7 Repeat step 3 through step 6 for the remaining subgrids.
Step 6. Save grid file
1 Click the Save Grid button on the main toolbar.
2 The grid is saved to a ImageName_grid.csv file where ImageName is the name of the image associated with the grid, and the directory containing the grid file is the same as the directory of the original image. A features file is saved at the same time as a ImageName_feat.csv file.
159
3 Creating Grid Files Manual Grid Adjustments
Manual Grid Adjustments
160
For most images, grids are adjusted adequately if you follow the steps provided by the Help Wizard. However, if the placement of the grid is not adequate after performing the steps of the Help Wizard, you can correct the grid placement using the following procedures.
To rotate the position of the main grid
To move or rotate the main grid
another way, see To move or
rotate the main grid with Grid
Adjustment pane on page 161.
1 Make sure that the color log scale has been turned on to view the corner spots on the main grid more easily.
2 Click the Adjust Main Grid button on the toolbar.
3 Move the mouse to a corner of the main grid in the lower Corner View.
The rotate cursor appears.
4 Drag the rotate cursor to rotate and optimize the position of the grid on the four corners of the microarray image.
Drag rotate cursor.
Fe
Creating Grid Files 3 To move or rotate the main grid with Grid Adjustment pane
Feature Extraction 12.2 User Guide
Figure 42 Final position of main grid on microarray
To move or rotate the main grid with Grid Adjustment pane
When the Main Grid View appears, the Grid Adjustment tab is the default selection. You can move or rotate the main grid by entering values for fields in the Grid Adjustment tab.
To learn all the ways that you can
move grids and spot centroids, see
Shortcuts to Help You Work with
Grids and Find Spots on page 177.
1 Click the Grid Adjustment tab, if necessary.
Figure 43 Grid Adjustment pane for Main Grid View
2 Click the Adjust Main Grid button on the toolbar.
3 Enter a value for any field that is available to you.
Array Label
Top Left Corner
Rotation
161
3 Creating Grid Files To rotate the subgrid
To rotate the subgrid
162
If moving the position of the subgrid does not work, try rotating the subgrid.
1 Click to select the Adjust Subgrid button on the toolbar.
2 Click to deselect the Select Corners and Automatically Fit Grid button on the toolbar (if it is turned on).
3 Click on a subgrid in the upper Main Grid View.
4 In the lower subgrid Corner View, move the mouse to a corner of the subgrid until the rotate cursor appears.
5 Drag the rotate cursor on the subgrid Corner View to optimize the position of the grid on the four corners of that portion of the microarray image.
Figure 44 Subgrid with rotate cursor
Rotate Cursor
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To skew the subgrid corner
To skew the subgrid corner
Feature Extraction 12.2 User Guide
If neither of the above methods work, try skewing the subgrid corner.
1 Click the Skew Subgrid button.
2 Move the mouse to a corner of the subgrid Corner View.
The skew cursor appears.
3 Drag the skew cursor on the subgrid Corner View to optimize the position of the grid on the four corners of that portion of the microarray image.
Figure 45 Subgrid with skew cursor
Skew Cursor
To position subgrids with Grid Adjustment pane
When the Adjust Subgrid button is selected, the Grid Adjustment tab is the default selection. You can position subgrids by entering values for fields in the Grid Adjustment tab.
163
3 Creating Grid Files To position subgrids with Grid Adjustment pane
To learn all the ways that you can
move grids and spot centroids, see
Shortcuts to Help You Work with
Grids and Find Spots on page 177.
164
1 Click the Grid Adjustment tab, if necessary.
Figure 46 Grid Adjustment pane for subgrids
2 Enter a value for any field that is available to you.
Col and Row Numbers
Click the down arrows for Col and Row to select the column and row numbers for the subgrid you want to position.
Corner of Subgrid
Rotation
Skew
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Saving a Grid File
Saving a Grid File
Feature Extraction 12.2 User Guide
Whenever you have made manual adjustments to the grid or spacing geometry, save the grid file in order to keep the changes you made. There is only one grid file for any given image.
1 Click the Save Grid button on the main toolbar.
2 The grid is saved to a ImageName_grid.csv file where ImageName is the name of the image associated with the grid, and the directory containing the grid file is the same as the directory of the original image. For non- Agilent images, a features file is saved at the same time, with the name ImageName_feat.csv.
165
3 Creating Grid Files Calculating Spot Size and Spot Centroids
Calculating Spot Size and Spot Centroids
This option is available only for
non-Agilent images.
166
This is the same algorithm used for calculating spot size and finding spot centroids by the program during an automatic extraction. For third- party arrays, this calculation shows you how well the algorithm will find the spots on the array and let you perform corrections, if necessary (see Locating Spot Centroids Manually on page 167).
1 Click the Adjust Spot button to enable spot mode.
2 Click the Calculate Spot Size and Centroids button to let you see the difference between the nominal spot centroid, positioned from the grid (+), and the estimated spot centroid based on the spot size (X).
After you have calculated the spot size and spot centroids, you can adjust manually. See Locating Spot Centroids Manually on page 167. When you move a grid or subgrid after calculating the spot size and spot centroids, the found centroids for the moved grid or subgrid are deleted.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Locating Spot Centroids Manually
Locating Spot Centroids Manually
If you want to save the manually
placed spots, save the grid file.
Feature Extraction 12.2 User Guide
Once you have calculated the spot sizes and centroids (see Calculating Spot Size and Spot Centroids on page 166) you can manually adjust them. In this section you learn how to move a nominal spot centroid, placed on a spot by the subgrid. When you move the nominal spot centroid (+), it becomes an estimated spot centroid (X). Once the spot centroid is estimated, the program uses this estimated value for Feature Extraction.
You cannot move spot centroids for
Agilent images. This function
works only for non-Agilent images.
If a spot centroid has been manually moved, the result called IsManualFlag is set to true. See the Feature Extraction 12.2 Reference Guide for the result.
167
3 Creating Grid Files To find a spot to adjust
To find a spot to adjust
168
1 Click the Adjust Spot button, to bring up the Main Spot View and the Zoomed Spot View.
Figure 47 Adjust spot view
Main View
Zoomed Spot View
Preview Spot Centroids button
Spot Navigator (square box)
Spot Navigator button
Spot Location Adjustment Pane
Adjust Spot button
The calculated spot centroids
appear when you enter Adjust Spot
mode.
2 Select Color > Use Log Color Scale if it has not already been selected.
3 Click a spot on the Main Spot View or on the Zoomed Spot View.
Or
Feature
Creating Grid Files 3 To find a spot to adjust
Feature Extraction 12.2 User Guide
Click the Spot Navigator button, to bring up the Grid Spot Navigator dialog box.
4 Click the down arrow in the Find by list, and select the number or name of the spot to navigate to.
5 In the Query field, type the feature number or probe name that you want to find.
Case is important.
6 Click Find.
If only one feature in the microarray matches the query, the program moves the spot square (Spot Navigator) to that feature.
If more than one feature matches, a list of features is displayed in the bottom box.
7 Select a feature from the list, and click Go to.
The program moves the Spot Navigator to that feature.
8 Click Close.
The program positions the Spot Navigator (square box) in the center of the Zoomed Spot View.
169
3 Creating Grid Files To position the spot centroid
To position the spot centroid
170
Before you position the Spot Navigator to find the centroid, the position of the nominal spot centroid could look like this:
Nominal spot centroid (+) to move
Move the Spot Navigator square containing the nominal spot centroid position to the assumed center of the spot.
The resulting position might look like this:
New estimated spot centroid (X) position;
These images represent non- Agilent microarrays. It is probably necessary to position the spot centroid for non- Agilent microarrays.
When this altered grid file is used in Feature Extraction, the default spot finding algorithm does not attempt to move spots that have been manually located in the grid file.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To ignore spots
To ignore spots
Feature Extraction 12.2 User Guide
1 Position the Spot Navigator where you want to ignore a spot.
2 In the Grid Adjustment pane, click the field next to Ignore and then select True.
The program does not use this spot
for any step in Feature Extraction.
To undo or redo spot position movements
Click the Undo button, to undo an action.
Click the Redo button, to redo an action.
171
3 Creating Grid Files To position spot centroids with the Grid Adjustment pane
To position spot centroids with the Grid Adjustment pane
172
When the Spot View appears, the Grid Adjustment tab is the default selection. You can position spot centroids by entering values for fields in the Grid Adjustment tab.
1 Click the Grid Adjustment tab, if necessary.
To learn all the ways that you can
move grids and spot centroids, see
Shortcuts to Help You Work with
Grids and Find Spots on page 177.
Figure 48 Spot Location Adjustment pane
2 Enter a value into any field that is available to you.
3 Press Enter.
Col and Row Numbers
Click the down arrows for Col and Row to select the column and row numbers for the spot centroid you want to position.
Center
SpotDiam
Feature Extraction 12.2 User Guide
Creating Grid Files 3 To position spot centroids with the Grid Adjustment pane
ProbeName
Feature Extraction 12.2 User Guide
Name of probe associated with selected spot; found in gene list
GeneName
Ignore
173
3 Creating Grid Files Refitting a Grid
Refitting a Grid
174
Refit a grid under the circumstances mentioned in the following table. Alternate solutions to the problems are also described:
Table 5 Conditions for refitting grids
Grid issue Things to try before refitting the grid
Main grid is not properly placed. Adjust the placement of the main grid. See
Step 3. Adjust the position of the main
grid on page 156.
Subgrids are not correctly spaced. Adjust the spacing between subgrids.
Assumption is that the subgrids are evenly
spaced. See Step 2. Get nominal subgrid
spacing manually on page 155 and Step
4. Adjust the subgrid on page 157.
Spot spacing appears off. Adjust the spacing between the spots.
Step 1. Get nominal spot spacing
manually on page 153.
Average spot size is not correct. Adjust the spot diameter. This diameter is
used by the Calculate Spot Size and
Centroids algorithm and the WholeSpot
algorithm. See Calculating Spot Size and
Spot Centroids on page 166 and
Locating Spot Centroids Manually on
page 167.
When you refit a grid, any changes you made during manual gridding, if
To refit a grid, follow these steps.
To learn about Geometry settings,
see Geometry information on
page 148.
1 Select Edit > Refit Grid.
The Geometry Information dialog box appears.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Refitting a Grid
Feature Extraction 12.2 User Guide
Click Reset to re-enter the original
number before you changed it.
Figure 49 Geometry Information dialog box to refit Agilent grid
2 Change entries, and click Auto Fit or Manual Fit. A message appears reminding you that refitting the grid causes you to lose any changes you made that you did not save in the grid file.
3 Click Yes to continue, or No to return to the Geometry Information dialog.
4 Once the grid is refit, save the grid file to keep the changes made by the refitting procedure.
See Saving a Grid File on page 165.
175
3 Creating Grid Files Changing Default Settings for Grid Editing
Changing Default Settings for Grid Editing
176
Before you begin to set up a grid, you can change the default settings that appear in the Grid Definition and Grid Adjustment panels on the main window when the grid mode is turned on.
1 Select Tools > Preferences.
2 Click the Grid Editing folder.
Figure 50 Preference SettingsGrid Editing selected
Default View Zoom setting
Click the cell next to an image window, then click the down arrow to select the zoom setting for each of the image windows listed.
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Shortcuts to Help You Work with Grids and Find Spots
Shortcuts to Help You Work with Grids and Find Spots
Feature Extraction 12.2 User Guide
There are several ways that you can select and move grids and spots. These options are presented in Table 6. In addition, there are a number of Ctrl and Alt key operations that help you work with grids. These options are presented in Table 7.
Table 6 Grid Actions
Location of actions
If you want to: Do this action on the Grid Adjustment dialog box:
Or, do this action with the keyboard:
Or, do this action with the mouse:
Main Grid Set array label Edit value, then press Enter. NA NA
Move grid Edit value, then press Enter. Hold down the Shift key
and select an arrow key.
Click the left mouse button
and drag the cursor across
the full window. Or, click the
left mouse button in the
corner window.
Rotate grid Edit value, then press Enter. NA Drag on corner in corner
window.
Subgrids Select subgrid* Select column and row
number from a choice list.
Use arrow keys or Page
Up/Page Down (requires
two rows or more) or
Home/End
Left-click mouse in full view
or corner view.
Move subgrid Edit location, then press Enter. Hold down the Shift key
and select an arrow key.
Drag the cursor across the
full window. Or, click the left
mouse button on the top left
spot of the subgrid in the
corner window.
Rotate subgrid Edit value, then press Enter. NA Drag on corner in corner
window. Or, turn on
automatic skew and rotate
and then click the left mouse
button on the upper left,
upper right, and lower left
spot of the subgrid in the
corner window.
177
3 Creating Grid Files Shortcuts to Help You Work with Grids and Find Spots
Skew subgrid Edit value, then press Enter. NA Click Skew Subgrid button,
then drag on corner in corner
window. Or, turn on
automatic skew and rotate
and then click the left mouse
button on the upper left,
upper right, and lower left
spot of the subgrid in the
corner window.
Spots Select spot* Select column and row
number from a choice list.
Use arrow keys or Page
Up/Page Down or
Home/End
Left-click mouse in full view
or corner view.
Move spot
centroid
Edit location, then press Enter. Hold down the Shift key
and select an arrow key.
Drag Spot Locator to position
spot centroid.
Flag spot Mark the Ignore this spot check box, then press Enter.
NA Double-click the right mouse
button to toggle this flag.
* These are the only grid actions that you cannot undo, redo or reset.
Table 6 Grid Actions (continued)
Location of actions
If you want to: Do this action on the Grid Adjustment dialog box:
Or, do this action with the keyboard:
Or, do this action with the mouse:
178
Table 7 Hot Key Actions for Grid Mode
If you want to do this: Press this hot key:
Turn Grid Mode On/Off Alt-G
Select for Main Grid adjustment Alt-M
Select for Subgrid adjustment Alt-S
Select for Spot adjustment Alt-P
Select to ignore a spot Ctrl-G
Undo Ctrl-Z
Undo All Ctrl-A
Redo Ctrl-Y
Feature Extraction 12.2 User Guide
Creating Grid Files 3 Shortcuts to Help You Work with Grids and Find Spots
Feature Extraction 12.2 User Guide
Redo All Ctrl-D
Refit grid Ctrl-T
Table 7 Hot Key Actions for Grid Mode (continued)
If you want to do this: Press this hot key:
179
180
3 Creating Grid Files Shortcuts to Help You Work with Grids and Find Spots
Feature Extraction 12.2 User Guide
Agilent Feature Extraction 12.2 User Guide
4 Changing Protocol Settings
Changing Protocol Settings 182
Creating a Protocol From an Existing Protocol 184
Changing Protocol Step 1: Place Grid 193
Changing Protocol Step 2: Optimize Grid Fit 200
Changing Protocol Step 3: Find Spots 204
Changing Protocol Step 4: Flag Outliers 213
Changing Protocol Step 5: Compute Bkgd, Bias and Error 220
Changing Protocol Step 6: Correct Dye Biases 237
Changing Protocol Step 7: Compute Ratios 243
Changing Protocol Step 8: MicroRNA Analysis 248
Changing Protocol Step 9: Calculate Metrics 252
Changing Protocol Step10: Generate Results 255
This chapter gives you background information on changing settings for the protocols used to perform Feature Extraction. The Feature Extraction protocols contain the parameter values for the algorithms that you specify for each step of the Feature Extraction process.
See Chapter 1 in the Feature Extraction 12.2 Reference Guide for a listing of the default settings and parameter values for the protocols shipped with the program.
181Agilent Technologies
4 Changing Protocol Settings Changing Protocol Settings
Changing Protocol Settings
Protocol templates
182
Agilent ships protocol templates for running Feature Extraction with various types of scan data. The name of each protocol template follows the format:
Agilent microarray type_program version_date
The table below lists the names of the templates that Agilent ships with Feature Extraction version 12.2, along with a description of changes from previous version. For a complete list of the default protocol settings, see the Feature Extraction 12.2 Reference Guide.
Both the protocol parameters and default values may change over time. To ensure you have the current protocol, from the Feature Extraction menu bar, click Help > Feature Extraction on the Web > Feature Extraction Protocols.
Table 8 Unremovable Protocol Templates
Protocol Template Name Description
CGH_1201_Sep17 Recommended for all Agilent CGH microarrays.
Updated from v. 11.5 to include:
The LogRatioImbalance metric, which
calculates the amount of amplifications versus
deletions per chromosome to determine if
there is an imbalance that falls out of normal
expectations.
ChIP_1200_Jun14 Recommended for all Agilent ChIP microarrays.
GE1_1200_Jun14 Recommended for all Agilent 1-color gene
expression microarrays using the latest wet lab
protocols.
GE2_1200_Dec17 Recommended for all Agilent 2-color gene
expression microarrays using the latest wet lab
protocols.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To create a new protocol from a protocol template
Feature Extraction 12.2 User Guide
Both the protocol parameters and default values change occasionally. To ensure that you have the latest protocols, set up eArray login and let Feature Extraction automatically check the eArray server for the protocol updates every time the program starts. See Setting up eArray Login for Automatic Updates on page 36.
GE2-nonAT_1100_Jul11 This is the recommended protocol for non-Agilent
microarrays.
miRNA_1200_Jun14 Recommended for all Agilent miRNA microarrays
using the latest wet lab protocols.
Table 8 Unremovable Protocol Templates (continued)
Protocol Template Name Description
To create a new protocol from a protocol template
See Chapter 5, How Algorithms
Calculate Results in the Feature
Extraction 12.2 Reference Guide for
detailed information on how the
protocol steps and algorithms
work.
Each protocol contains protocol steps that represent Feature Extraction algorithms. In the FE Protocol Editor, you make changes to the parameter values for each of the steps.
After you open the FE Protocol Editor for a specific protocol, you can change settings, then save the protocol to a new name. The process is as follows:
1 To open a protocol.
2 To view or change the protocol properties.
3 To change the protocol steps to run.
4 To change the settings for each protocol step.
5 To save a protocol with a new name.
The following sections present instructions for creating protocols from templates and tasks for changing settings.
183
4 Changing Protocol Settings Creating a Protocol From an Existing Protocol
Creating a Protocol From an Existing Protocol
184
To create a protocol for a specific type of microarray, use an existing protocol for the same type of microarray as a template.
For example, if you want to create a new protocol to run a gene expression 2- color microarray, start by editing an Agilent- created protocol for gene expression 2- color microarrays.
You can track what the original Agilent- created protocol is by looking at the Derived From field of the general section of the new protocol. See To change the protocol steps to run on page 187.
CAUTION Existing protocols can be used as templates for new protocols to
provide both visible and hidden settings whose values are specific to
the type of microarray. Although you can change the visible settings of
any two protocols of different types to appear identical, you cannot change the hidden settings that distinguish these protocols from one
another.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Creating a Protocol From an Existing Protocol
To open a protocol
Feature Extraction 12.2 User Guide
You can open the FE Protocol Editor in one of three ways:
In the Protocol Browser double- click the protocol that you intend to edit, or right- click the protocol and select Properties.
In Project Explorer, either double- click the protocol assigned to an extraction set, or right- click the protocol and select Properties.
Click the Extraction Configuration tab, and in the Protocol column double- click the protocol that you intend to edit.
Regardless of the protocol, the first screen that appears is by default the screen that was last displayed before a protocol was closed or saved.
Figure 51 FE Protocol Editor Protocol Steps
185
4 Changing Protocol Settings Creating a Protocol From an Existing Protocol
To view or change the protocol properties
186
The General section of the Protocol Properties page lists fields to identify the protocol and fields to let you remove the protocol or make the protocol read- only.
If necessary, in the FE Protocol Editor click the Protocol Properties button to see the General section.
Figure 52 FE Protocol Editor Protocol Properties tab
Date and Time
Protocol Version
Derived From
Metricset
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change the protocol steps to run
Protocol Type
Feature Extraction 12.2 User Guide
Corresponds to type of microarray the protocol is used for in Feature Extraction: 2- color gene expression, 1- color gene expression, CGH, ChIP, or miRNA data.
Protocol Removable
Describes if the protocol is removable from the database. User- created protocols are removable. Agilent- created protocols are not removable.
Description
Permanent Read Only
True indicates that the protocol cannot be edited. Save any changes made to the protocol with a different name.
False indicates that you can change the protocol. You can save any to the protocol with the existing name.
To change the protocol steps to run
If there is no protocol on your system with the steps you want Feature Extraction to use, you can open a protocol of the appropriate type and change the protocol steps.
1 If necessary, in the FE Protocol Editor click the Protocol Properties button to see a list of the protocol steps on the right.
2 Mark each check box to enable a step that you want to run. Clear each check box to disable a step that you do not want to run.
Refer to Figure 52 to see the FE Protocol Editor.
187
188
4 Changing Protocol Settings To change the protocol steps to run
Protocol steps for GE 2-color, CGH, and ChIP images
Below is a description of each of the protocol steps for GE 2- color, CGH, and ChIP images, as seen in Figure 52, in the order that the program runs them.
Place Grid
For any microarray loaded with a grid file that you saved during Grid Mode (i.e. interactive gridding), automatically calculates centroid positions. Any spots that were manually adjusted are left unchanged.
For Agilent microarrays, the program performs a background peak shift that helps find peaks of features that are close to the background. See the Feature Extraction 12.2 Reference Guide for details.
For any microarray, user can select to use regions near to the center of packs for slope and skew calculations. This option is of use in cases where pack edges have dim spots and are failing to grid. When this is selected, the user can select to have the program to perform one extra step of correlation following the projection data analysis to get the origin.
Optimize Grid Fit
Find Spots
From the located spots, the features are defined with either a CookieCutter method or a WholeSpot method.
The local background is also defined with a radius method.
Outlier pixels (too low or high intensity) are then removed from the pixel populations that defined the features and backgrounds.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change the protocol steps to run
Feature Extraction 12.2 User Guide
Statistics, such as the mean, median and standard deviation (SD), are calculated on the remaining inlier pixels within the feature and background.
Features with a large number of saturated pixels are flagged.
The protocol determines whether the mean or median signals are used for further analysis.
Flag Outliers
Features that have replicates of the same probe sequence and background regions are flagged as population outliers if their signal distributions are outside the set range.
Compute Bkgd, Bias, and Error
If background subtraction is selected, background signal is subtracted from raw signal for each feature using a local or global background method.
A statistical test is performed to assess if the feature signal is significantly different from the background signal. This test uses either the error model or pixel statistics, as indicated in the protocol choice Significance.
Raw signals or background- subtracted signals can be corrected for systematic gradients, that affect low signals especially, by using additive detrending (i.e. spatial detrending).
The error is calculated at this point, using the choice between the Universal error model or the larger of the error estimates (i.e. most conservative) from the Universal error model and the propagated error.
A multiplicative detrending algorithm is also used to correct for gradients proportional to the signal.
Surrogates, based upon background noise, can be used as processed signals for features that have low- intensities, near the background noise level.
Correct Dye Biases
The method for how a dye normalization gene list is selected is set.
189
190
4 Changing Protocol Settings To change the protocol steps to run
Differences in the red and green channel signals caused by different efficiencies in labeling, emission and detection are estimated and corrected.
Features to be used for measuring dye bias are selected based on a probe selection method chosen.
A dye normalization curve fit method is selected and used to fit the data for measuring dye bias. (SNP probes are not included in dye normalization calculations.)
Compute Ratios
Calculate Metrics
In this step, the Stats table is generated and QC metrics are calculated for the QC Report.
To generate a complete Stats table and QC report, leave this option turned on. If this option is turned off, some values are not included.
The type of protocol you select determines the type of QC report that is generated.
For example, a CGH protocol generates a CGH QC Report, and a 1- color protocol with spike- ins selected generates a 1- color QC Report with spike- ins.
Generate Results
Results selected for the project are generated.
In the Protocol Editor you can select to generate one file for the three tables (FEPARAMS, STATS, FEATURES) of text results or three separate files. You can set the compression for the jpeg file.
Protocol steps for GE 1-color and miRNA images
GE 1-color Feature Extraction analyzes GE 1- color images using the same steps listed for GE 2- color images except it does not use the Correct Dye Biases and Compute Ratios steps.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change the settings for each protocol step
Feature Extraction 12.2 User Guide
miRNA The miRNA analysis also does not correct for dye biases nor compute ratios. Instead, the analysis estimates the expression level by sampling multiple probes with multiple features per probe and reporting a total gene signal (counts/pixel), independent of microarray design, rather than a processed signal.
The protocol, then, does not contain the Correct Dye Biases step nor the Compute Ratios step, but it does contain a MicroRNA Analysis step before the Calculate Metrics step.
To change the settings for each protocol step
After you change the protocol steps that you want to run, you may want to change the default settings and parameter values for a specific protocol step.
1 Click the Protocol Steps tab to see a list of the protocol steps, or Feature Extraction algorithms, on the left.
2 Click the protocol step button whose settings you intend to change.
Figure 53 FE Protocol Editor with Protocol Steps on left
191
192
4 Changing Protocol Settings To save a protocol with a new name
See the remaining sections in this chapter to learn how and why you change parameters for each protocol step.
To save a protocol with a new name
After you make any change to a protocol, save the protocol to a new name.
1 Make the changes you intend.
2 Click Save As.
Figure 54 Add new protocol dialog box
3 Enter a Protocol name.
4 If you want to make the protocol editable, clear the check box Save as permanent read only.
For standard protocols that you use frequently, you may want to save these protocols as permanent, read- only protocols so that others cannot alter them.
5 Click OK.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step 1: Place Grid
Changing Protocol Step 1: Place Grid
Feature Extraction 12.2 User Guide
The Place Grid algorithm initiates Feature Extraction by placing a grid on the microarray image. The grid comes from either the grid template or the grid file associated with the image file in the project.
To access the settings to place a grid
Click the Place Grid tab in the FE Protocol Editor, if not already there.
Figure 55 FE Protocol Editor Place Grid selected
193
4 Changing Protocol Settings To access the settings to place a grid
To select the microarray format
194
This topic describes how to select an array format if needed. You typically do not need to do this step, as the program automatically determines which array format and scan type is used.
Once you change to a specific array format, the protocol loses its ability to be general and only works for that specific format.
1 Click the cell to the right of Array Format.
2 Click the down arrow, and select a microarray format from the list.
Figure 56 Place Grid Array Format list
The Placement Method option now appears. See Figure 57 on page 195.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To access the settings to place a grid
Feature Extraction 12.2 User Guide
Figure 57 Place Grid 30-micron multi pack Array Format selected,
Placement Method
After you select the format, you can select a method for placing the grid.
Protocols can automatically detect the array format (65- micron feature size or 30- micron feature size). The protocol automatically makes different parameter choices depending on the format and type. These choices are transparent to you and are primarily in the image processing (Place Grid and Find Spots).
With this feature, you do not need different protocols for different array formats. These protocols contain the correct parameter choices for each and use the set that matches the format.
195
196
4 Changing Protocol Settings To select a method for placing the grid
Information about which parameters are affected and which parameters are chosen for each format type can be found beneath any option with a choice automatically determine. If you believe an incorrect automatic choice was made, you can select a specific format. However, this is not a normal or necessary step.
To select a method for placing the grid
1 Click the cell next to Placement Method, and click the down arrow.
2 Select one of the two options listed below.
Allow Some Distortion
Allows some variance in the feature spacing, subgrid spacing and the skew. All of the protocol templates have this default setting.
Place and Rotate Only
Only allows rotation of the grid and no changes to feature spacing, subgrid spacing or the skew.
To select to use enhanced gridding
Feature Extraction 12.2 offers an enhanced gridding algorithm for CGH protocols that was not available in protocols that were created prior to version 12.2. These enhancements were designed to improve automatic gridding in array images of lower quality, including those with non- linear features. Array images that may have previously received a failed grid error and require manual gridding can now be processed successfully using the enhanced algorithm.
Among the default protocols for Feature Extraction 12.2, only the CGH protocol (CGH_1201_Sep17) uses enhanced gridding. Agilent has not validated and does not currently
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select to enable background peak shifting
Feature Extraction 12.2 User Guide
recommend the use of enhanced gridding in GE1, GE2, ChIP, or miRNA protocols. If you are editing an imported protocol of any type, the enhanced gridding setting is off by default.
1 Click the cell next to Use Enhanced Gridding, and click the down arrow.
2 Select True or False.
If you have been using an older
version of Feature Extraction (prior
to v12.1) for CGH microarrays and
want to start using enhanced
gridding, Agilent recommends
validating the new enhanced
gridding algorithm before
incorporating into your standard
procedures. Once validated, you
can use the new default CGH
protocol (CGH_1201_Sep17), or
import your previously used
protocol and set the enhanced
gridding feature to True.
When set to True, the enhanced gridding algorithm that was added in Feature Extraction 12.1 is used in the protocol.
When set to False, the enhanced gridding algorithm that was added in Feature Extraction 12.1 is NOT used in the protocol. The previous version of the gridding algorithm is used instead.
To select to enable background peak shifting
1 Click the cell next to Enable background peak shifting? and click the down arrow.
2 Select True or False.
If you select True, the background pixel value is obtained using the histogram of the image. All the pixels in the image that belong to background are reassigned a value of zero. Note that this assignment is temporary and does not change the actual image data for subsequent steps.
197
4 Changing Protocol Settings To select to use central part of pack for slope and skew calculations
To select to use central part of pack for slope and skew calculations
198
This option lets you choose the region to use for slope and skew calculation during subgrid placement.
1 Click the cell next to Use central part of pack for slope and skew calculations? and then click the down arrow.
2 Select True or False.
When set to True, regions near to the center of packs are used. This option is of use particularly in cases where pack edges have dim spots and are failing to grid. (This is the default setting for 30 micron single- and multi- pack array formats.)
When set to False, the regions near to the edges of packs are used. Selecting this option uses the same parameters as Feature Extraction version 10.7 or earlier.
To select to use the correlation method to obtain origin x of subgrids
This option lets you choose the technique to use to determine X location of origin of subgrids.
1 Click the cell next to Use central part of pack for slope and skew calculations? and click the down arrow.
2 Select True.
3 Click the cell next to Use the correlation method to obtain origin x of subgrids? and then click the down arrow.
4 Select True or False.
When set to True, the program performs one extra step of correlation following the projection data analysis to get the origin. This option is of use particularly in cases where pack edges have dim spots and are failing to grid.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select to use the correlation method to obtain origin x of subgrids
Feature Extraction 12.2 User Guide
When set to False, the results obtained from the projection data analysis are used to estimate the origin. This is how the origin is estimated using Feature Extraction version 10.7 or earlier.
199
4 Changing Protocol Settings Changing Protocol Step 2: Optimize Grid Fit
Changing Protocol Step 2: Optimize Grid Fit
200
The Optimize Grid Fit step is a post grid placement process to improve the grid registration. Leveraging from the spot finder algorithm, this step examines the spots in the four corners of the microarray to iteratively adjusts the grid for best fit.
To access the settings to optimize the grid fit
Click the Optimize Grid Fit tab in the FE Protocol Editor.
Figure 58 FE Protocol Editor Optimize the Grid Fit selected
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To access the settings to optimize the grid fit
To select the grid format
Feature Extraction 12.2 User Guide
1 Click on the cell next to Grid Format.
2 Click the down arrow, and select from the list of grid formats.
Figure 59 Optimize Grid Fit Grid Format list
If you select either 65- micron or 30- micron feature size, the Iteratively Adjust Corners? option appears. See Figure 60 on page 202.
201
202
4 Changing Protocol Settings To change the settings to iteratively adjust corners when grid format is selected
Figure 60 Optimize Grid Fit 30-micron Feature Size selected,
Iteratively Adjust Corners
To change the settings to iteratively adjust corners when grid format is selected
1 Click the cell next to Iteratively Adjust Corners?.
2 Either select True or False.
3 If you select True, you can change any of the default parameters that you need to.
When set to True, the Interactively Adjust Corners method runs the spot finder algorithm in the corners to see how well the grid is placed on the microarray and iteratively adjusts the grid based upon the results.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change the settings to iteratively adjust corners when grid format is selected
Adjustment Threshold
Feature Extraction 12.2 User Guide
The grid is adjusted based upon the difference between the grid position and the spot position if the absolute average difference is greater than this value.
Maximum Number of Iterations
This is the maximum number of times the spot finder algorithm runs to adjust the grid for better grid fitting in the corner spots.
Found Spot Threshold
Number of Corner Features Side Dimension?
This specifies the number of features (i.e. spots) on each side of the square in each corner of the microarray to be used to calculate the average difference.
The default number is 20 for the 30- and 65- micron feature size microarrays. This means each corner of the microarray has a square area of 400 (i.e. 20x20) features that are calculated.
203
4 Changing Protocol Settings Changing Protocol Step 3: Find Spots
Changing Protocol Step 3: Find Spots
See Chapter 1, Default Protocol
Settings, of the Feature Extraction
12.2 Reference Guide for a setting
comparison between the various
microarray formats.
204
After Feature Extraction places the grid, it is ready to locate the spot centers and define the features and local background regions for each spot. The options available and default values for each setting depend on the microarray format selected in Step 1: Place Grid.
To access the settings to find spots when spot format is automatically determined
Click the Find Spots tab in the FE Protocol Editor.
Figure 61 FE Protocol Editor Find Spots selected
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select the statistical method to reject pixel outliers
A spot is a circular or amorphous
area on the microarray containing a
hybridized sample.
A feature is defined as the pixels in
either the center portion of a spot
or in the whole spot, depending on
the method you choose to define
the feature.
Feature Extraction 12.2 User Guide
When the spot format is automatically determined (default setting), these are the only settings available for change:
Select the statistical method to reject outlier pixels from features and local background before Feature Extraction calculates feature and background statistics.
Select whether to use mean or median for data analysis
To select the statistical method to reject pixel outliers
1 Click the cell next to Pixel Outlier Rejection Method, and click the down arrow.
2 Select one of these three choices:
Standard Deviation
This calculation rejects outliers outside of a specified number of standard deviations from the mean of the intensity for all pixels. If selected, you can enter the number of standard deviations (Feature SD and Background SD).
Inter Quartile Region
This calculation rejects outliers based on the Interquartile Range (IQR). See the Feature Extraction 12.2 Reference Guide for how this calculation works. If selected, you can enter the IQR factor for both the feature and background (RejectIQRFeat and RejectIQRBG).
None
Feature Extraction rejects outlier pixels after calculating statistics on the intensities of all the pixels of a feature or local background region for each color. If a pixel is rejected in one color, it is rejected in both colors.
205
4 Changing Protocol Settings To select whether to use mean or median for data analysis
To select whether to use mean or median for data analysis
206
1 Click the cell next to Statistical Method for Spot Value from Pixels, and click the down arrow.
2 Select one of these choices to be used in the next step of data analysis:
Mean
Median
During the background correction stage the background can be subtracted from either the mean signal or the median signal. This option lets you decide which signal type to use for downstream analysis.
To enable the Enhance Spot Finding option (CGH protocols only)
1 Click the cell next to Use Enhanced SpotFinding, and click the down arrow.
2 Either select True or False.
If you select True, then the enhanced spot finding method for CGH arrays is used during the spot finding step of the protocol. This method allows for more accurate identification of the center of each spot and of the pixels that comprise each spot. The result is that fewer features are called as non- uniform.
If you select False, then the protocol does not use the enhanced spot finding method.
The Use Enhanced SpotFinding setting is disabled when the Use Enhanced Gridding setting (in the Place Grid menu) is set to False. See To select to use enhanced gridding on page 196.
Note that enabling the Use Enhanced SpotFinding setting can slightly increase the array processing time.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To enable the Enhance Spot Finding option (CGH protocols only)
To access the settings to find spots when spot format is selected
Feature Extraction 12.2 User Guide
1 Click the cell to the right of the Spot Format.
2 Click the down arrow, and select from the list of spot formats.
Figure 62 Find Spots 30-micron Spot Format selected
In this protocol step you can also change the values for 3 additional sets of parameters associated with Feature Extraction processes to find spots once you have selected a spot format:
Change values to find and size spots.
Select a spot statistics method to define features.
Select to autoestimate the local radius.
207
4 Changing Protocol Settings To change values to find and size the spots.
To change values to find and size the spots.
208
1 Click the cell next to the names of the following parameters.
2 Either select True or False, or enter a value, whichever is applicable.
Use the Nominal Diameter from the Grid Template
If you select True, the nominal diameter of each spot is obtained from the grid template. If you select False, the nominal diameter that is the starting spot size assumption for the Spot Finding algorithm is obtained from the grid placement algorithm.
Spot Deviation Limit
Enter a value that defines the distance the actual centroid may move relative to its nominal grid position (stated as a multiple of nominal radius). This parameter appears in the FEPARAMS table as SpotAnalysis_kmean_cen_reject.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change values to find and size the spots.
To select a spot statistics method to define features
Feature Extraction 12.2 User Guide
1 Select either Use Cookie or Use Whole Spot.
Cookie Method
The Cookie method is usually used
for Agilent microarray spots
because of their uniformity.
Figure 63 Feature and background definitions Cookie method
Local backgroundring area between blue
dashed circle and red dashed circle
Exclusion zonering area between feature
and blue dashed circle outside of spot. The
pixels in this zone are not included in feature
or local background statistics.
Edge of spotblack circle
Cookie or featuregreen circle in middle
Whole Spot Method
This method uses a detection technique based on a noise model to define the feature and exclusion zone. This method performs the spot size calculation first and defines the feature as the whole spot from the x and y diameters.
Figure 64 Feature and background definition Whole Spot method
Local backgroundarea between blue
dashed circle and red dashed circle
Exclusion zonearea between edge of
feature (spot) and blue dashed circle outside
of spot
Edge of spotblack contour
Featuregreen area; entire spot
209
210
4 Changing Protocol Settings To change values to find and size the spots.
2 If you selected Use Cookie, change the Cookie Percentage and/or the Exclusion Zone Percentage, if necessary.
If you selected Use Whole Spot, change the Exclusion Zone Percentage, if necessary.
Cookie Percentage
Exclusion Zone Percentage
Percentage of the nominal radius that is the exclusion zone radius.
Figure 65 Definitions of different radii
Exclusion zone radius from centroid of
spot to edge of blue-dashed line.
Cookie radius from center of
spot to edge of green circle
Nominal radius from spot centroid
to edge of spot
Local background radius from spot centroid to edge
of red circle
See the Feature Extraction 12.2
Reference Guide for more
information.
The radii depicted in Figure 65, Definitions of different radii, on page 210 are calculated differently for the two methods.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change values to find and size the spots.
To select to autoestimate the local radius to define local background
Feature Extraction 12.2 User Guide
This radius method estimates background that is local to a feature of interest.
1 Click the cell next to Autoestimate the local radius.
2 Select True or False.
For non-Agilent microarrays when
there is a big difference in spot
sizes, change this setting to False
and make sure that the number you
enter is big enough to
accommodate the largest spot on
the microarray.
3 If you select False, enter a larger radius if you want to include more background pixels or a smaller one if you want to include fewer.
If you select True, the program automatically estimates the radius to use to estimate the local background.
This approach eliminates any dependence of the local background computation on distance between spot- centroids along both axes. This ensures elimination of cross- contamination between feature and background signals, especially in closely packed microarrays.
See the Feature Extraction 12.2
Reference Guide for details on this
calculation.
The default radius produces a local background calculation around only one spot (Self- radius). If you want to include more background, you can increase the radius to include the background around the spots nearest neighbors (NN radius).
211
4 Changing Protocol Settings Propagation of Find Spot errors to other algorithms
Final calculations for protocol step to find and measure spots
212
Final statistics
The final statistics calculated for all colors of each feature and local background, based on the remaining inlier pixels, are the average, standard deviation, median, IQR, and number of pixels. In addition, the pixel correlation between the red and green channels is calculated.
Pixel saturation
Saturation is defined as the
maximum allowed intensity value
for a pixel, Its value depends on the
make and model of the scanner.
Pixel saturation is also tested during calculation of the feature statistics. For the Agilent Microarray Scanner system, saturation is defined as an intensity value of 65535 minus the dark offset from the scanner. If more than 50% of the pixels remaining after outlier rejection are saturated, Feature Extraction sets a saturation flag.
With XDR extractions, the saturation limit could be as much as 20x higher than 65535. The array will saturate before it gets to 20x of 65535. The calculated saturation value based upon the XDR fit can be as high as 1,310,000 counts, but no biological signals are that high.
20- bit scans also have a high saturation value, one close to 1,048,576 counts.
Propagation of Find Spot errors to other algorithms
The measurement errors from the calculation of these statistics are propagated to the next steps in Feature Extraction according to standard error propagation techniques. For example, the errors for the signal average for the feature and local background regions are propagated to the background subtraction algorithm.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step 4: Flag Outliers
Changing Protocol Step 4: Flag Outliers
Feature Extraction 12.2 User Guide
Not every feature is created, hybridized and scanned perfectly. Some backgrounds contain artifacts. Therefore, you can flag those features or local backgrounds that may represent misleading or erroneous data.
Although the grid tool of the Feature Extraction program lets you manually examine scanned microarray images and flag anomalous features, this task becomes exceedingly tedious and subjective for thousands of features on a microarray. This algorithm automatically finds all the anomalous features on the microarray.
To access the settings to flag outliers
Click the Flag Outliers tab to change settings for flagging abnormal features and backgrounds.
Figure 66 FE Protocol Editor Flag Outliers selected
213
214
4 Changing Protocol Settings To change settings to flag population outliers
You can change two types of settings for this algorithm:
Change settings to flag population outliers.
With this option you can flag features and local background regions that have intensities beyond the confidence limit for the distribution of the intensities for the replicate features and local background regions on the microarray.
Change settings to flag non- uniform features and background region (outliers).
To change settings to flag population outliers
A type of outlier, called a population outlier, can be detected if there are enough replicates to perform population statistics on the signal intensities of the features or the signal intensities of the background regions.
For example, there are a number of control features on a typical Agilent microarray (typically, over 100 features) and thousands of local background regions. Out of this population, outliers can be rejected using box plot analysis.
1 Click the cell next to Compute Population Outliers, and click the downarrow.
2 Select True or False.
3 If you select True, change any of the default parameters that you need to.
The Flag Outliers tab contains values for the minimum population size to be considered for population outlier flagging, as well as a cutoff range. If a feature or local background region signal lies outside of this range, Feature Extraction flags that feature or local background region as a population outlier.
Minimum population
This is the minimum number of replicates needed to perform the population outlier analysis. Although 10 is the default number of replicates, you can enter as low as 5 replicates.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change settings to flag non-uniform outliers
Feature Extraction 12.2 User Guide
To take advantage of the replicated probes in statistical calculations, Agilent uses the default threshold of 10 because its catalog microarrays have 100 probes that are replicated 10 times on the microarray. The default value for non- Agilent microarrays is 15.
IQRatio
Background IQRatio
This option uses the IQRatio formula for estimating the background region outliers.
Use Qtest for Small Populations?
When the number of replicates falls below the minimum number specified in the protocol, a Q- test can be used to estimate the number of population outliers instead of the IQRatio. See Chapter 5 of the Feature Extraction 12.2 Reference Guide to learn more about Q- test calculations.
Report Population Outliers as
Failed in MAGEML file
When this option is set to True, any feature that is flagged as a population outlier is reported as Failed in the MAGE- ML output file.
To change settings to flag non-uniform outliers
The Agilent Microarray system
includes Agilent microarrays,
label/hyb/wash protocols and
reagents, scanner and Feature
Extraction algorithms.
The non- uniformity outlier algorithm flags anomalous features or local backgrounds based upon statistically significant deviations from a noise model that makes use of the known noise characteristics of the Agilent Microarray system.
You do not typically need to do this step, as the program automatically determines the scan type and sets the parameters appropriately. To view or change the parameters
215
216
4 Changing Protocol Settings To change settings to flag non-uniform outliers
that are automatically set, select a specific scan type. However, if you change the parameters, the protocol is no longer general and can only be used with that scan type.
1 Click the cell next to Compute NonUniform Outliers, and click the downarrow.
2 Select True or False.
True is the default setting for all Agilent- created protocols. If True is set, move on to the next step.
3 Click the cell next to Scanner, and click the down arrow.
4 Select either Automatically Determine or Agilent Scanner, NimbleGen Scanner, or Innopsys Scanner.
The settings change depending on your selection.
To learn more about how the
algorithm calculates the terms in
the polynomial equation below, see
the Feature Extraction 12.2
Reference Guide.
5 Click the cell next to Automatically Compute OL Polynomial Terms.
6 Select True or False.
True is default setting. A different set of parameters appears for a True or a False setting.
7 Enter the values for the parameters.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change settings to flag non-uniform outliers
Feature Extraction 12.2 User Guide
Parameters for a True setting
Figure 67 Flag Outliers Agilent Scanner selected,
Automatically Compute OL Polynomial Terms: True
The polynomial equation for the noise model is still used, but rather than entering values for the A, B and C terms as in the False case, you enter the values for the A (%CV)^2 term and multipliers for the B and C terms in both color channels (2- color) or one channel (1- color).
(%CV)2 Term (A)
Red Poissonian Noise Term (B) Multiplier
The B term for the red channel is equal to this multiplier times the net signal at the 25% percentile of a histogram plot of number of negative control features or background vs. net signal.
217
4 Changing Protocol Settings To change settings to flag non-uniform outliers
Red Signal Constant Term (C) Multiplier
218
The C term for the red channel is equal to this multiplier times the variance at the 25% percentile of a histogram plot of number of negative control features or background vs. variance.
Green Poissonian Noise Term (B) Multiplier
The B term for the green channel is equal to this multiplier times the net signal at the 25% percentile of a histogram plot of number of negative control features or background vs. net signal.
Green Signal Constant Term (C) Multiplier
The C term for the green channel is equal to this multiplier times the variance at the 25% percentile of a histogram plot of number of negative control features or background vs. variance.
CAUTION The default values for these terms were optimized for Agilent
microarrays, with an Agilent scanner and recommended lab protocols.
You perform optimization experiments to determine the optimum
values for your microarrays and lab protocols.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change settings to flag non-uniform outliers
Feature Extraction 12.2 User Guide
Parameters for a False setting
Figure 68 Flag Outliers Agilent Scanner selected,
Automatically Compute OL Polynomial Terms: False
A noise model, based upon a polynomial equation, is made up of three variables whose values are shown in the Automatically Compute OL Polynomial Terms folder:
(%CV)2 Term (A)
Poissonian Noise Term (B)
This term specifies the variance due to Poisson distributed noise, which accounts for error in counting statistics.
Background Term (C)
This term specifies the variance due to signal- independent background noise (e.g., scanner and glass substrate noise).
219
4 Changing Protocol Settings Changing Protocol Step 5: Compute Bkgd, Bias and Error
Changing Protocol Step 5: Compute Bkgd, Bias and Error
220
At this stage of the Feature Extraction process, several systematic sources may still remain that contribute to the background signal component of the feature signal:
Scanner offset, which can vary from 20 counts to 300 counts, depending upon the version of Agilent scanner [B (20) or A (300) scanner]. The value for each microarray offset is shown in the Stats table in the fields: gDarkOffsetAverage and rDarkOffsetAverage (See the Feature Extraction 12.2 Reference Guide.)
Glass and any contaminants or non- specifically- bound fluorescent signal on the glass. Artifacts from labeling, hybridization and washing steps can contribute to this source.
Fluorescent signal that is non- specifically associated with the DNA probes themselves.
A systematic gradient across a microarray.
Background region levels can also vary depending on the dye label and color channel used.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step 5: Compute Bkgd, Bias and Error
To access the settings to compute background, bias and error
Feature Extraction 12.2 User Guide
Click the Compute Bkgd, Bias and Error tab to select a method and change settings for background subtraction.
Figure 69 FE Protocol Editor Correct Bkgd, Bias and Error
The selections for this protocol step let you decrease the contribution of the background signal to the feature signal.
To view the values used for
background subtraction depending
on the subtraction and correction
methods selected, see the Feature
Extraction 12.2 Reference Guide.
Select the background subtraction method.
Choose to decrease the signal contribution due to an additive gradient at low signal across the microarray spatial detrend.
Choose to adjust the signal contribution due to a systematic multiplicative gradient across the microarray multiplicative detrend. SNP probes are not included in multiplicative detrending.
221
222
4 Changing Protocol Settings To select the background subtraction method
Choose to globally adjust the initial background- subtracted signals.
Choose an error method to estimate the error in the calculations for the processed signal.
Change settings to assess if the background- subtracted signal is significant compared to the background.
Choose to use or not use surrogates to improve the reliability of low- signal error calculations.
To select the background subtraction method
There are several methods for subtracting the background signal from the raw signal, both local and global, available in the Correct Bkgd and Signal Biases sheet.
1 Click the cell next to Background Subtraction Method, and click the down arrow.
2 Select the background subtraction method most relevant to your microarray and processing conditions.
The default method for Agilent microarrays is No background subtraction. The default method for non- Agilent microarrays loaded with a grid file is Minimum signal (feature) on the microarray. This option makes the fewest assumptions about the microarray system. If you are analyzing non- Agilent microarrays or procedures, you have to determine the optimum default method.
No Background Subtraction
Even though no background is subtracted, you can make corrections to the signal using spatial detrend and the global background adjustment tools. The spatial detrend algorithm corrects for sources of background that are available in this protocol step, in addition to correcting for spatial gradients.
Local background
method
This method subtracts the local background estimated for each spot using the radius method in the Find Spots protocol step. For example, local background for spot 300 is subtracted from the feature raw signal of spot 300.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select the background subtraction method
Feature Extraction 12.2 User Guide
The local background method may be a better background subtraction method than a global background subtraction method if your microarray has large regions of local backgrounds with much higher signals than the rest of the microarray.
These regions are often due to hybridization or wash processing artifacts. The local background method subtracts these local variations. This yields a better estimate of true feature signal if the local background signal is correlated with the feature signal, as determined by replicate statistics.
Global methods The Feature Extraction program supports a number of global background subtraction methods in case local background subtraction is not appropriate. With a global method, one number is calculated as background for the entire array. That number is subtracted from every feature. The number is reported in the STATS table for later queries. Global methods require that both non- uniformity outlier and population outlier methods be turned on. See the topic Changing Protocol Step 4: Flag Outliers on page 213 for more information.
Average of All Background
Areas
This global method subtracts the average of the local background inliers estimated for all the spots using the radius method in the Find Spots protocol step. Inlier background regions have passed both the non- uniformity and population algorithms.
For example, the average background for all spots is subtracted from the feature raw signal of spot 300 and every other spot on that microarray.
Average of Negative Control
Features
If the local background is not additive and your microarray has validated negative controls, Agilent recommends using the Average of negative control features option. If the local background is not additive and your microarray does not have negative controls, Agilent recommends the Minimum signal (feature) option.
223
4 Changing Protocol Settings To select the background subtraction method
Negative controls are probes that
consistently show very little target
binding across many types of
targets and have no specific target
bound to them by design.
224
This global method subtracts the average of the inliers of all the negative control features from each feature on the microarray. Inlier negative control features have passed both the non- uniformity and population algorithms. For example, the average of the inlier negative control features is subtracted from spot 300 and every other spot on that microarray.
Agilent provides validated negative
controls for catalog microarrays.
These probes have been shown to
reflect the non-specific signal level
that is present on probes.
If there is more than one probe sequence on the microarray annotated in the grid template or grid file as a negative control, this method analyzes the population of feature signals of each sequence to determine if there are significant differences among the different sequences at a fixed significance level of 0.01. The method then chooses the probe sequences that have the lowest mean signals and that are not significantly different from each other as the population of features to calculate negative control statistics.
Minimum signal (feature
or background)
This global method subtracts the minimum signal on a microarray from each spot, whether the signal is a feature or local background region. The feature or background region signal chosen for the minimum signal must pass a low pixel standard deviation requirement. For example, the minimum feature or background signal on the microarray is subtracted from spot 300 and every other spot on that microarray.
Minimum signal (feature)
on the array
This global method subtracts the minimum signal of a microarray feature that has passed a low pixel standard deviation requirement. This setting is the default setting for microarrays loaded with a grid file, but not necessarily the recommended setting. For example, the minimum feature signal on the microarray is subtracted from spot 300 and every other spot on that microarray.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To choose to remove the surface trend found in the data
To choose to remove the surface trend found in the data
Feature Extraction 12.2 User Guide
What is detrending?
To learn more about the detrending
calculations, see the Feature
Extraction 12.2 Reference Guide.
Removing the surface trend (detrending) decreases the contribution to the background of any systematic signal gradient on the microarray. For both 2- color and 1- color data, Feature Extraction removes the additive surface trend from the data, a process called spatial detrending (background detrending). Feature Extraction also removes any domed surface trend, a process called multiplicative detrending. SNP probes are not included in multiplicative detrending calculations.
This option is the default choice with protocols for Agilent microarrays (except the miRNA protocol). The detrending is done individually for each color channel and corrected separately.
Background detrending consists of two steps: calculating a surface fit and subtracting the surface fit from the data. This process is done independently for the red and green channels after background subtraction and before the global background adjustment.
Background Spatial Detrending
For spatial detrending, Agilent separates the surface fit and subtraction calculations that must exist together to remove the surface trend found in the data.
The foreground is the portion of the
signal in a feature that is not
related to the intended signal from
the dye-labeled target
complementary to the probes on
that feature.
The surface fit calculation estimates the foreground of the data by measuring the intensity of the feature after background subtraction (or intensity of the MeanSignal of the feature if the No background subtraction option is selected) and performs a surface fit to the selected data. Even if the surface fit is not subtracted from the data (perform spatial detrending), the calculation appears in the QC Report to help evaluate your data.
225
226
4 Changing Protocol Settings To choose to remove the surface trend found in the data
The advantage of this method is that it can remove variation in foreground intensity for different regions of the microarray. By performing spatial detrending and subtracting the surface fit through this foreground, the data becomes more consistent and reproducible.
Feature Set for Surface Fit
AllFeatureTypes Select if microarray has a set of very low intensity features evenly distributed on the slide.
OnlyNegativeControlFeatures Select this option if you have enough negative controls that are evenly spread around the microarray and ideally have different sequences represented default for CGH and ChIP protocols.
FeaturesInNegativeControlRange Select for low intensity features that are within 3 standard deviations of the spatial interpolated values of the negative controls - default for gene expression and miRNA protocols.
Perform Filtering for Surface Fit
If True, allows a subset of the features into the set to be used to calculate the surface fits. You may want to do this filtering to reduce the time needed to calculate the surface without reducing the accuracy of the fit.
If the Feature Set is AllFeatureTypes, then the filter chooses 1 to 2 percent of the dimmest features to be selected across the array. Select True for AllFeatureTypes. Select False for OnlyNegativeControlFeatures.
If the feature Set is FeaturesInNegativeControlRange, the filter chooses 1 to 2 percent of the features at random to go into the surface fitting set. Random x% rather than dimmest x% is used because the features in this set are already known to be dim.
Perform Spatial Detrending
If true, subtracts any additive trend found from the surface fit from the data.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To choose to remove the surface trend found in the data
To choose to adjust the background globally (2-color)
Feature Extraction 12.2 User Guide
To correct the errors at the low end of the signal range, Agilent lets you choose to globally adjust the background after initial background subtraction. The global adjust method was developed to be used by those customers who decide to turn off all background and detrending methods.
This adjustment helps estimate and remove background signal due to fluorescent signal that is non- specifically associated with the DNA probes and that may be under- estimated by upstream background subtraction methods. You can also select this correction if you also chose No background subtraction.
1 Click the cell next to Adjust Background Globally, and click the down arrow.
2 Select True or False.
3 If you select True, enter a pad value between 0 and 500 to force the log ratio towards zero when both red and green signals are small.
The constant supplied to the Adjust background globally algorithm pads all the feature signals by that value. This option has the effect of compressing log ratios, but decreases the variability (standard deviation) in the log ratio among inter- and intra- array replicates. You can try small pad values, typically in the few tens, and monitor the trade- off between log ratio compression and log ratio variability. The log ratio versus signal plot is a useful preliminary tool to assess the effect of the pad value.
By setting the Adjust background globally algorithm to the same constant with multiple microarrays, the residual background can now be defined by you and is therefore more consistent across multiple microarrays and multiple experiments than if a variable residual background signal is left for each feature, if no background is subtracted.
227
4 Changing Protocol Settings To choose to remove the surface trend found in the data
To choose to perform multiplicative detrending
228
Multiplicative detrending is an algorithm designed to compensate for slight linear variations in intensity if the processing is not homogeneous across the slide, resulting in different chemical reaction concentrations, for example, between the sides and the center, a dome effect. The multiplicative gradient is removed by dividing the signal by the observed gradient. SNP probes are not included in multiplicative detrending calculations.
Make the selections described below.
Perform Multiplicative
Detrending
Choose whether to perform multiplicative detrending (can be chosen for one- or 2- color arrays).
Detrend on Replicates Only
Choose whether to detrend on all of the data or to only use those probes that have replicates. If you choose to use replicates only, the data are normalized to the respective replicate averages before going into the fitting set.
Filter Low signal probes
from Fit?
You can chose to filter out dim signals from the fit of the multiplicative trend surface. This option is useful because the dimmer signals have a less obvious trend when that trend is multiplicative. You also choose how dim the signals must be in order to be removed from the set. This value is stated as a multiple of the Negative Control Spread.
Perform filtering for fit
You decide whether to use all of the data points or use local averages or medians.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select the method to estimate the error (Error Model)
To choose to use robust negative control statistics
Feature Extraction 12.2 User Guide
This algorithm was developed to be used with microarrays that have many different probes classified as negative controls, such as CGH. The population outlier analysis, described on the previous pages, is performed on these negative controls, one probe sequence at a time.
Select a Robust Neg Ctrl Stats? option.
True
Once outliers are identified, they are excluded from the calculation of the robust negative control statistics (average and standard deviation).
False
See the Feature Extraction 12.2 Reference Guide for more information on how this algorithm works.
To select the method to estimate the error (Error Model)
1 Click the cell next to Choose universal error, or the most conservative (propagated or universal), and click thedownarrow.
2 Select an error method from one of these options:
Universal Error
Using this error model, a p- value is calculated that corresponds to the probability that the difference between the red and green channels is zero. The Universal Error Model is a good estimator of error in high intensity features.
229
230
4 Changing Protocol Settings To select the method to estimate the error (Error Model)
It may underestimate the error of low signal features or features that have low signal- to- noise (e.g., non- uniform features).
Most Conservative
Propagated Error This method measures the error on the log ratio by propagating the pixel- level error of that feature through all the calculations on the raw signal, including raw pixel statistics, background subtraction, dye normalization and the log ratio calculation. Agilent follows standard propagation of error techniques and calculates a p- value that corresponds to the probability that the log ratio = 0.
The propagated error is a good representation of noise that predominates at low signal ranges, such as instrumentation noise and the noise due to non- uniform features.
The propagated error cannot capture other sources of noise that dominate at higher signal ranges, such as biological and chemical variability (e.g., writing and labeling, hybridization and wash processes). The propagated error method usually underestimates the noise, especially at higher signal levels.
See Parameters for Universal and Most Conservative error models on page 230for more instructions on changing the parameters for these two error models.
Parameters for Universal and Most Conservative error models
Parameters to change With this error model and the most conservative error model, you can enter a change to the following settings and parameters:
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select the method to estimate the error (Error Model)
MultErrorGreen (Red)
Feature Extraction 12.2 User Guide
Enter the multiplicative error component in the channel selected
Autoestimate Additive Error
Green(Red)
You can select to enter your own constant for the additive error or turn on auto- estimation of the additive portion of the error model. If you select True, then Feature Extraction auto- estimates the additive error. If you select False, then enter the constant.
The constant setting of the additive error approximates the error of feature signals that are near background. If the overall signal of the microarrays is increased or decreased (e.g. due to hotter or cooler labeling or variable scanner PMT settings), the additive error needs to be adjusted. The auto- estimation method of determining the additive error adjusts the scale accordingly.
CAUTION The default additive and multiplicative parameters have been
optimized using the most current Agilent microarray-based system
(e.g., microarrays, protocols, scanner and most recent Feature
Extraction algorithms). Important: If you use non-Agilent microarrays
or Agilent microarrays with non-Agilent protocols, re-evaluate these
values to reflect the noise in your system.
Auto-estimation of additive noise Provided that the microarray, platform and protocols remain the same, the additive error can vary from one microarray to the next. Therefore, it is useful to correctly estimate the additive noise and set the additive constant equal to this noise.
To do this the program looks at microarray statistics: the width of the negative control distribution used in the background surface fit and the spatial variability of the spatial detrend surface (RMS of the difference between each point on the surface and the mean of the surface). Because the algorithm uses these surface statistics in the calculation, this option is only usable for microarrays for which surface calculation works.
231
232
4 Changing Protocol Settings To select the method to estimate the error (Error Model)
You do not have to turn on Spatial detrending for the additive error to be automatically estimated. You do have to turn on the surface fit calculation for the additive error to be estimated.
Reliability of auto-estimation of additive error Agilent has found that the auto- estimation of the additive error results in error values that are more reliable than those generated if a constant is used. If the microarray quality or processing is less than optimal, the additive error is more reliable with auto- estimation. This makes the significance calls more correct and less susceptible to errors induced by microarray and processing variability.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select the method to estimate the error (Error Model)
To choose to use or not use surrogates
Feature Extraction 12.2 User Guide
Agilent Feature Extraction program uses surrogate signals to calculate a more accurate and reproducible log ratio for signals of low intensity.
You choose to use surrogates in this protocol step so that the signals to be used for log ratios are subjected to dye normalization to remove this kind of bias.
Agilent Feature Extraction program
uses the error estimate option that
you select for all features, no
matter if surrogates are used for
the log ratio calculations for some
features.
1 Click the cell next to Use Surrogates, and click the down arrow.
2 Select True or False.
Conditions under which surrogates are used
If you select True, the program uses surrogates under two conditions:
If either red or green MeanSignal fails the IsPosAndSignif test (i.e., IsPosAndSignif = 0)
If BGSubSignal is less than the Error
A log ratio cannot be calculated from a negative ratio. For any measurement system there is increased confidence in using a signal only if it is greater than a lower limit that is particular to that system. This ensures that the system lower limit is enforced as a lower limit on the signal.
Feature Extraction calculates the surrogate signal from the Error. For all default Agilent protocols that use Error Model significance, this is simply the Additive Error calculated during background subtraction.
233
4 Changing Protocol Settings To change settings to test the significance of a feature
Propagation of errors from the Compute Bkgd, Bias and Error algorithm to other algorithms
234
The measurement errors from the background subtraction are propagated to the next steps in Feature Extraction according to standard error propagation techniques. For example, the errors for the feature and background signals are propagated with the error on the background- subtracted signal to the DyeNorm algorithm.
Estimation of error is performed in this step. See To select the method to estimate the error (Error Model) on page 229 to see the choices show for this step. In this case, you cannot select the propagated error calculation by itself. If you select Most Conservative, Feature Extraction uses the larger of the two estimates from the Universal Error model or the propagated error.
To change settings to test the significance of a feature
Determining if a signal is significant, or just in the noise of the background, is an important step. The following two tests are performed to establish the significance of a feature: a two- sided t- test and a method to flag features that are well above background.
2-sided t-test of feature vs. background
maximum p-value
The t- statistic is calculated based on the feature and background signals and their corresponding errors.
The background value used for this test depends upon the background method used. The values are reported in the Feature Extraction text output column g(r)BGUsed. If you have chosen not to subtract the background and Spatial detrend = Off, the local background for each feature {g(r)BGMeanSignal} is used for the t- test but is not subtracted from the feature signal. If Spatial detrend = On when the background subtraction is off, the surface value for each feature is used for the t- test and is subtracted from the feature signal.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To change settings to test the significance of a feature
To view the values used for the
t-test depending on the subtraction
and correction methods selected,
see the Feature Extraction 12.2
Reference Guide.
Feature Extraction 12.2 User Guide
Estimating background error for significance testing uses one of two methods. The method used is dependent upon the protocol parameter, Significance (for IsPosAndSignif and IsWellAboveBG).
Pixel significance
For this method, the SD of the background that is used for the t- test is the same value used for surrogates (See Use of surrogates for log ratio and p- value calculations on page 244 of this guide). This is a population level SD for global background methods and is a pixel- level SD for both the local background and No background methods. These SD's are reported in the Feature Extraction text output columns g(r)BGSDUsed. (Not used for error model significance, which is the favored method.)
The number of degrees of freedom is also based on the error calculations and the number of pixels, which depends on the background subtraction method. The p- value is then calculated based on a two- sided t- test. If this p- value is less than the maximum p- value defined in the protocol, the feature is considered to be significantly different from the background.
Note that entering a positive pad
value for global background adjust
may increase the number of
features considered positive and
significant in the t-test, in some
cases, all features.
The features that fail the null hypothesis, that is, have a signal that is significantly different from the background and also that have background- subtracted signals > 0, have a boolean set = 1 in the Feature Extraction text output column g(r)IsPosAndSignif. The features that are not significantly different from background or that have background- subtracted signals < 0 have the boolean set = 0 for the same Feature Extraction text output column.
WellAboveMulti
235
236
4 Changing Protocol Settings To change settings to test the significance of a feature
Error model significance
The other method to determine feature significance is based on the error model. The protocol parameter Significance (for IsPosAndSignif and IsWellAboveBG) is set to Use Error Model for Significance for this method. The additive error is used as the estimate of background noise and is used for surrogates.
The significance test for positive and significant then becomes the probability of the feature actually having zero signal for a feature with a given background- subtracted signal with a known additive error (or noise of features with no true signal) assumed to be gaussian distributed about zero.
You choose a p- value to decide how far along the Gaussian distribution the feature needs to be before it is considered Significant. When this choice is made, the Well Above Background is decided by a multiple of the additive error.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step 6: Correct Dye Biases
Changing Protocol Step 6: Correct Dye Biases
A dye bias is the difference in the
red and green signals caused by
different efficiencies in labeling,
emission or detection.
Feature Extraction 12.2 User Guide
Systematic differences in dye bias must be removed from the 2- color GE, CGH, and ChIP data. To remove dye bias, the Feature Extraction program uses a dye normalization algorithm. This step is not used with GE 1- color and miRNA data.
To access the settings to correct dye biases
Click the Correct Dye Biases button to change settings to remove difference in dye bias.
Figure 70 FE Protocol Editor Correct for Dye Biases selected
You can change the following types of settings for this algorithm.
Select how dye normalization gene list is determined
Select features that are included in the normalization set
Select to omit background population outliers
Select to allow positive and negative controls
237
238
4 Changing Protocol Settings To choose how to select dye normalization list
Select the signal characteristics
Select a curve to determine the bias from that normalization set of features
The program normalizes the entire microarray from this measured bias.
If you use a self versus self experiment, in which the feature signals should be the same in both channels across the dynamic range of signal intensities, differences in dye behavior can result in the signals in one channel increasing across the signal range at a different rate than in the other channel.
This behavior can be visualized by plotting the background- subtracted signals of one channel versus the other and noting that the slope of the line fit through the data points is not equal to one. Dye normalization corrects for this by multiplying the signals by appropriate dye normalization factors.
To choose how to select dye normalization list
Three options are available for choice of the dye normalization gene list for correction of dye biases.
1 Click the cell next to Use Dye Normalization Gene List, and click the downarrow.
2 Select one of the following choices.
Automatically Determine
The DyeNormList is used, if present in the grid template. Otherwise, the list is generated internally, based on the Dye Normalization Probe Selection Method. If no DyeNormList is selected in the grid template, extraction will not fail.
True
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To choose a method to select a set of dye normalization probes
False
Feature Extraction 12.2 User Guide
When selected, a DyeNormList is not used even if it is specified in the grid template. During the dye normalization step, an internal dye normalization list is generated, based on the Dye Normalization Probe Selection Method.
To choose a method to select a set of dye normalization probes
Feature Extraction supplies two methods for selecting representative features that constitute the set used for bias measurement.
1 Click the cell next to Dye Normalization Probe Selection Method, and click the downarrow.
2 Select one of the following methods.
Use Rank Consistent
Probes
This is the recommended method of selecting normalization features. The rank consistency filter selects features that fall within the central tendency of the data by observing consistent trends between the red and green channels. This method assumes that the central tendency of the data is unregulated. This method requires that the features pass the three criteria described below:
They have not been flagged as non- uniform in either channel.
They are not saturated.
The condition for saturation is more stringent in this algorithm (> 5% of pixels saturated) than in the algorithm for features as saturated (> 50% of pixels saturated).
They are not population outliers in either channel.
Rank Tolerance
239
240
4 Changing Protocol Settings To select to omit background population outliers
To assign a red rank and a green rank, order all of the probes based on the intensity of their background- subtracted signal. A rank 1 is the lowest intensity signal feature. The highest rank is assigned to the most intense signal. A perfectly rank consistent feature would be one exactly at the 75% percentile of the distribution in both the red and green signal.
Variable Rank Tolerance
You can choose to allow variable rank tolerance. This does rank tolerance after grouping the data into signal groups. In each group the same designated percentage of most rank- consistent data are included.
Use All Probes
To select to omit background population outliers
1 Click the cell next to Omit Background Population Outliers, and click the down arrow.
2 Select True or False.
Selecting True excludes any features from the dye normalization set if the local backgrounds associated with those features have been flagged as population outliers (in either channel). The default recommendation is False.
To select to allow positive and negative controls
If you turn this function on, Feature Extraction uses positive and negative controls in the dye normalization calculation.
1 Click the cell next to Allow Positive and Negative Controls?, and click the down arrow.
2 Select True or False.
The default recommendation is False.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select signal characteristics
To select signal characteristics
Feature Extraction 12.2 User Guide
You can select to include features with the following signal characteristics as part of the set of rank consistent features.
1 Click the cell next to Signal Characteristics, and click the down arrow.
2 Select one of the following options.
OnlyPositiveandSignificant Signals
Use only positive and significant signals for ranking in the rank consistency filter.
AllPositiveSignals
AllNegativeAndPositiveSignals
To select a normalization correction method
After you select a set of normalization features, you select a normalization method to apply to that set. The results from the calculation are applied to all features on the microarray.
1 Click the cell next to Normalization Correction Method, and click the down arrow.
2 Select from one of the following methods:
Max Number of Ranked Probes
The maximum number of ranked probes allows only a certain number of probes into the dye normalization set. If the number of rank- consistent probes exceed this number, a random set is chosen to meet the maximum number. If the number entered is - 1, this filter is not used. This option can speed up the data processing without adversely affecting accuracy for data sets with large numbers of probes, especially for 65- micron spot size and higher density arrays.
Linear
241
4 Changing Protocol Settings To select a normalization correction method
Linear and LOWESS
242
This is the default method for normalizing Agilent microarrays. Dye biases tend to be signal intensity- dependent. A dye normalization method that uses a constant scaling factor cannot correct for dye biases across the entire range of signal intensities. This method does a linear normalization across the entire range of data, then applies a non- linear normalization (see LOWESS only description below) to the linearized data set.
LOWESS only
A well- known curve fitting method called LOWESS (a locally weighted linear regression curve fit) is used to fit this curve. Given any feature and its intensity, this curve determines the potential dye bias that should be present in the log ratio measurement of the feature. Features are then corrected by multiplying the red and green channels by an adjustment that removes the bias from the log ratio.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step 7: Compute Ratios
Changing Protocol Step 7: Compute Ratios
Log ratio is the log (base 10) of the
ratio of the red channel signal to
the green channel signal.
Log ratio error is the error about
the mean of the log ratio.
P-value is a measure of confidence
that a feature is, or is not,
differentially expressed.
Feature Extraction 12.2 User Guide
To determine if a feature is differentially expressed, three measures are important: log ratio, log ratio error and the p- value.
The log ratio error is determined by the error model and therefore used to help assess the significance of the computed log ratio.
The p- value is calculated using the log ratio and the log ratio error. Genes with lower p- values have higher confidence of being differentially expressed. For example, if the p- value is <= 0.01, there is a 99% confidence level that the gene is differentially expressed.
This step is not used with GE 1- color and miRNA data.
To access the settings to compute ratios
Click the Compute Ratios tab in the FE Protocol Editor to change the settings to calculate these measures.
See Figure 71 on page 244.
243
244
4 Changing Protocol Settings To enter the value for the log ratio limit
Figure 71 FE Protocol Editor Compute Ratios selected
You can change only one parameter for this protocol step:
Enter the value for the log ratio limit.
To enter the value for the log ratio limit
When you set a limit, the features whose log ratio exceeds that limit, either in a positive or negative direction, are not used in any calculations. The recommended limit for this value is 4.
1 Click the down arrow in cell next to Peg Log Ratio Value.
2 Enter a positive or negative value up to 4.
Use of surrogates for log ratio and p-value calculations
Surrogates are used in this step to calculate certain log ratios, p- values and log ratio errors.
The table below shows the conditions for use of surrogates to calculate log ratios, log ratio errors and p- values.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Use of surrogates for log ratio and p-value calculations
*For log ratios equal to or between -4 (ratio = .0001) and +4 (ratio = 10000). If (r/G) > 1 or (R/g) < 1, log ratio = 0 and p-value = 1.
Table 9 Use of surrogates for log ratio and p-value calculations
If red channel signal is And green channel signal is Then ratio is
And Feature Extraction does this*.
Positive and significant vs.
background and greater than
red background SD (R)
Positive and significant vs.
background and greater than
green background SD (G)
R/G Uses the dye-normalized signals for both
channels for the calculations.
Not positive and significant vs.
background or less than red
background SD (r)
Positive and significant vs.
background and greater than
green background SD (G)
r/G Uses the dye-normalized green signal and
the dye-normalized red background SD signal
(surrogate) for the calculations.
Not positive and significant vs.
background or less than red
background SD (r)
Not positive and significant vs.
background or less than green
background SD (g)
r/g Uses surrogates for both channels.
Sets the log ratio = 0.
Sets the p-value = 1.
Positive and significant vs.
background and greater than
red background SD (R)
Not positive and significant vs.
background or less than green
background SD (g)
R/g Uses the dye-normalized red signal and the
dye-normalized green background SD signal
(surrogate) for the calculations.
Table 10 Summary Use of surrogates for calculations
Case 1: R/G
Both channels use DyeNorm Signals.
P-value and log ratio are calculated as usual.
For signals not using surrogates,
g(r)DyeNormSignal = g(r)ProcessedSignal,
which is then used to calculate log ratio.
Case 2: r/G
r = rSurrogateUsed
G = gDyeNormSignal
P-value and log ratio are calculated as usual.
If r/G > 1, then Feature Extraction automatically sets
LogRatio = 0 and PvalueLogRatio = 1
Case 3: R/g
R = DyeNormSignal
g = gSurrogateUsed
P-value and log ratio are calculated as usual.
If R/g < 1, then Feature Extraction automatically sets
LogRatio = 0 and pValueLogRatio = 1
Case 4: r/g
Both channels use surrogates.
Feature Extraction automatically sets
LogRatio = 0 and pValueLogRatio = 1
For signals using surrogates,
g(r)ProcessedSignal =
g(r)SurrogateUsed x g(r)DyeNormFactors.
Feature Extraction 12.2 User Guide 245
246
4 Changing Protocol Settings Use of surrogates for log ratio and p-value calculations
Effect of using surrogates on log ratios
Below are plots that show the difference between using surrogates and not using surrogates for a set of self vs. self microarray data.
Figure 72 Plot of LogRatio vs. Avg_grProcessedSignalNo surrogates
Green stars are data points affected by use of surrogates in Figure 73.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Use of surrogates for log ratio and p-value calculations
Feature Extraction 12.2 User Guide
Figure 73 Plot of LogRatio vs. Avg_grProcessedSignalWith surrogates
Green stars are data points affected by use of surrogates. See Figure 72 for their positions without surrogates.
Effect of using surrogates on log ratio errors
If Use Surrogates is True, the program calculates log ratio errors using the dye- normalized and/or surrogate signals, according to the error model selected for the calculation.
If Use Surrogates is False, the program calculates log ratio errors using the dye- normalized signals, according to the error model selected for the calculations. If either or both channels have negative dye- normalized signals, the p- value is set to 1 and the log ratio error is set to 1000. If red channel < 0, then Feature Extraction sets the log ratio = - 4. If green channel < 0, then Feature Extraction sets the log ratio = +4. If both red and green channels < 0, then Feature Extraction sets the log ratio = 0.
247
4 Changing Protocol Settings Changing Protocol Step 8: MicroRNA Analysis
Changing Protocol Step 8: MicroRNA Analysis
248
This step appears only in the miRNA protocol. With this step you set up to calculate a Total Gene Signal. See the Feature Extraction 12.2 Reference Guide for more details on how Feature Extraction processes a miRNA image and on the QC report and text file results, algorithms and protocol default values.
Both the protocol parameters and default values may change over time. To ensure that you have the latest protocols, set up eArray login and let Feature Extraction automatically check the eArray server for the protocol updates every time the program starts. See Setting up eArray Login for Automatic Updates on page 36.
To access the settings to analyze microRNA (miRNA) images
Click the MicroRNA Analysis tab in FE Protocol Editor to change the settings for miRNA analysis.
Figure 74 FE Protocol Editor miRNA protocol microRNA Analysis
selected
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To choose to generate the results to a GeneView file
Feature Extraction 12.2 User Guide
You can change the following settings for this protocol step.
Choose to generate the results to a GeneView file, or not.
Choose to scale the analysis with the effective feature size parameters, or not.
Enter an Is Probe Detected multiplier.
Choose to exclude the non detected probes, or not.
Choose to automatically determine feature size fraction or by array type.
To choose to generate the results to a GeneView file
After running Feature Extraction with the microRNA protocol, a new file is generated in addition to the tab- text file, MAGE- ML file, QC Report and any other files selected. This file is saved in the same directory as the other result files and is named ArrayName_GeneView.txt.
Each row includes a gene on the microarray, or a unique Systematic/Gene name in the design. This file is a tab- text file with the following columns:
SystematicName
ControlType
gTotalGeneSignal
gTotalGeneError
IsGeneDetected
1 Click the cell to the right of Output GeneView File.
2 Click the down arrow, and select True or False.
249
4 Changing Protocol Settings To choose to scale the analysis with the effective feature size parameters
To choose to scale the analysis with the effective feature size parameters
250
The calculation of the feature size is different for miRNA images. Even though you can change the default value from True to False, you almost never have to do this.
1 Click the cell to the right of Analyze By Effective Feat Size.
2 Click the down arrow, and select True or False.
Even though the default value is True and these parameters appear, you almost never have to change them:
Maximum Number of Features
Maximum number of features used to compute the effective size.
Minimum Number of Ratios
Minimum number of ratios used to compute the effective size.
Low signal Percentile
Signal filter for computing the effective size of the spot. Uses signal in at least n (default50%) percentile of all signal.
High Signal Percentile
Signal filter for computing the effective size of the spot. Uses signal in at most n (default90%) percentile of all signals.
Minimum Noise Multiplier
Makes sure the signals are at least x times above the background noise for use in computing the effective spot size.
Throw Away Ratios Greater
Than
Ratios > 1 mean the outside of the cookie is brighter than the inside of the cookie. This may be due to crescents or non- uniformity or other anomalies.
To change the Is Probe Detected multiplier
This determines the Is Probe Detected multiplier in the miRNA algorithm.
1 Click the cell next to Is Probe Detected Multiplier.
2 Enter the desired value from 1.0 to 20.0. (The default value is 3.00.)
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To choose to exclude non detected probes
To choose to exclude non detected probes
If this option is set to False, the
results generated are the same as
in Feature Extraction 10.5.
Feature Extraction 12.2 User Guide
This changes how the TotalGeneSignal is calculated. If a TotalProbeSignal is not detected, then it is not added to the Total Gene Signal. In the miRNA analysis, the features relating to each probe is first turned into Total Probe Signals. The TotalProbeSignals for each probe associated with the target miRNA (Gene) are then summed. If a probe associated with an miRNA is not detected because it fails the IsProbeDetected flag, then if this option is True, it will not contribute to the TotalGeneSignal, and its error will not propagate to the TotalGeneError.
Set the Total Gene Signal to
the Total Signal Error
By default this is set to False. When this option is set to True, the Default Total Gene Signal if all genes are not detected option is not used. This option is used only if all the probes in an miRNA (Gene) are not detected and the non detected probes are excluded from the calculation (Exclude Non Detected Probes set to True.) If this option is set to True, the value of the TotalGeneError if all probes were included is used for the TotalGeneSignal. Another option appears that allows the user to set a fixed value for the TotalGeneSignal if no probes are detected. That value by default is 0.10.
Default Total Gene Signal if all
probes are not detected
Used only if all the probes in an miRNA are not detected, and the non detected probes are excluded from the calculation (Exclude Non Detected Probes set to True.) For an miRNA where all of its probes are not detected, then the TotalGeneSignal is set to this value to avoid errors caused by the algorithm trying to calculate a log of 0.
251
4 Changing Protocol Settings Changing Protocol Step 9: Calculate Metrics
Changing Protocol Step 9: Calculate Metrics
252
In this step you can change the settings for the metrics that appear in the QC report.
To access the settings to calculate metrics
Click the Calculate Metrics tab in the FE Protocol Editor.
Figure 75 FE Protocol Editor Calculate Metrics selected
You can change the following parameters in this step:
Select True for Spikein Target Used if Agilent spike- in targets were used during lab processing of the microarray. The appropriate metrics are generated in the QC report.
Enter the minimum population for replicate statistics.
Select the grid test format.
Enter the p- value for differential expression.
Enter the percentile value.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 To select to use spikein targets
To select to use spikein targets
Feature Extraction 12.2 User Guide
1 Click the cell next to Spikein Target Used?, and click the down arrow.
2 Select True or False.
If you select True, Feature Extraction does all the calculations necessary to report the statistics for both the non- control features and the spikein controls.
To change the minimum population for replicate statistics
The QC report includes statistics based on replicate probes (%CV, log ratio uniformity). Feature Extraction requires the minimum population of replicates in order to calculate these statistics.
1 Click the cell next to Minimum Population for Replicate Statistics.
2 Enter a number of 3 or greater.
To select the grid test format
1 Click the cell next to Grid Test Format.
2 Select the grid format of the images to be tested, or select Automatically Determine to let the program decide for you.
253
4 Changing Protocol Settings To change the p-value for differential expression
To change the p-value for differential expression
254
The QC report contains a plot of the log ratio vs. the magnitude for each feature to show which features are up- regulated, which are down- regulated and which show no differential expression. This assessment is relative to the p- value chosen for this setting.
1 Click the cell next to PValue for Differential Expression.
2 Enter a p- value against which you intend to measure differential expression on the microarray.
To change the percentile value
This value is an indicator of the distribution of signals of the non- control features and is useful in downstream normalization of 1- color data.
1 Click the cell next to Percentile Value.
2 Enter a value to set the distribution of signals of the non- control features.
Feature Extraction 12.2 User Guide
Changing Protocol Settings 4 Changing Protocol Step10: Generate Results
Changing Protocol Step10: Generate Results
To access the settings to generate results
Feature Extraction 12.2 User Guide
Click the Generate Results tab in FE Protocol Editor to change the values for output.
Figure 76 FE Protocol Editor Generate Results selected
You can change the following settings for this protocol step:
Select the QC Report type.
Select to generate a single file for the text output.
Change the compression factor for the JPEG output.
To select the type of QC Report
1 Click the cell next to Type of QC Report, and click the down arrow.
2 From the list, select the QC report type to be generated.
255
4 Changing Protocol Settings To select to generate a single file for the text output
Gene Expression
256
Generate a Gene Expression QC report. For more information, see the Feature Extraction 12.2 Reference Guide.
Streamlined CGH
Generate a Streamlined CGH QC report. For more information, see the Feature Extraction 12.2 Reference Guide.
CGH_ChIP
miRNA
Test Chip
To select to generate a single file for the text output
1 Click the cell next to Generate Single Text File?, and click the down arrow.
2 Select True or False.
If you select true, the text output appears in a single file. If you select False, it appears in three separate files, one for FEPARAMS, one for STATS and one for FEATURES (results).
To change the compression factor for the JPEG output
The value for this factor determines the degree to which the image is scaled down in both directions before being converted to JPEG.
1 Click the cell next to JPEG Down Sample Factor, and click thedownarrow.
2 Enter a number.
Feature Extraction 12.2 User Guide
Agilent Feature Extraction 12.2 User Guide
5 Changing Image Displays
Displaying and Saving Image Files 258
Loading Feature Extraction Visual Results 264
Changing Image Display Settings 271
Changing Feature Extraction Window Displays 293
This chapter presents instructions on changing image displays and working with image files to view the images before and after extraction.
257Agilent Technologies
5 Changing Image Displays Displaying and Saving Image Files
Displaying and Saving Image Files
To open an image file
258
1 Click on the toolbar Open image file button, or click File > Open > Image, to display the Open Dialog Box.
2 In the Open Dialog Box, double- click a .tif file from the scanner, or select a file and click Open.
You can also open an image file by
dragging the file to the Feature
Extraction desktop icon.
An image of the scanned microarray contained in the file appears in the Workspace.
Note that for images produced on a NimbleGen, Molecular Devices, or Innopsys scanner, the program first converts the image before opening it.
Figure 77 Uncropped TIFF image for Agilent microarray
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To crop an image
Feature Extraction 12.2 User Guide
You can also open an image file by double- clicking the image file in a project extraction set. For information on how to do open an image file, see Chapter 2, Extracting Microarrays Automatically.
To crop an image
Crop an image using the cursor
You do not have to crop an Agilent image before you load it into a project and extract. Crop images to view a subset of features or to use for illustration purposes.
When you first open an image, the
Cropping Mode button is turned
ON because it is the default
setting. (See To change default
image viewing settings on
page 287.)
1 If not already on, click the Cropping Mode On/Off button.
2 Position the crop cursor at the top left corner of the image that you want to crop, and hold the left mouse button down.
3 Drag the crop cursor to the opposite corner.
4 Release the left mouse button to display the cropped image.
Figure 78 Image top menu and toolbar in Feature Extraction Tool
Cropping mode ON
The Status Bar is located at bottom
of the Feature Extraction window
and contains the pixel information
about the viewed image. If the
viewed image is the preview
image, the pixels are from the
preview. If the viewed image is the
high-resolution image, the pixels
are the high-resolution pixels.
If you are viewing high- resolution images, crop them to one- tenth or less (that is, at least 200%) of their original size in order to see the features at high resolution. Any larger cropped image is a preview image.
The color of the crop cursor helps you recognize the proper size for the high fidelity image. If the cursor is black, the cropped image is the high fidelity image. If the cursor is red, the cropped image is still a preview image.
259
260
5 Changing Image Displays To crop an image
You can cancel the cropping operation by using the (Esc) key, or click the Cropping Mode On/Off button to turn cropping off.
Crop an image using corner locations (manual crop)
This option lets you crop an image with greater precision than you can attain with the cursor option.
NOTE For manual cropping to work, the image after cropping must be a minimum
of 30 x 30.
1 Click Edit > Manual Crop.
Figure 79 Manual Crop dialog box
2 Enter the locations of the corners (diagonal to one another) that you intend to have as the corners of the cropped image.
3 Click OK.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To crop an image
To save a cropped image to a new file
Feature Extraction 12.2 User Guide
If the image file is from an Agilent microarray with a barcode:
1 Click File > Save Modified Image to save the cropped image to a new file.
Figure 80 Save Modified Image dialog box for Agilent microarray
2 To close the file after saving, clear the Open file after saving check box.
3 To specify which pack in the multiplex (formerly known as multipack) microarray was cropped, select a Multiplex_Row_Column number.
This option lets you save the cropped single pack with its multiplex coordinates.
If no multiplex was cropped, clear the Cropped Multiplex check box.
4 To change the New Array Identifier, see To add or change the New Array Identifier on page 263.
For Agilent microarrays with a barcode, the New Array Identifier is automatically generated from the barcode and displayed, along with a suggested file name.
261
262
5 Changing Image Displays To crop an image
The New Array Identifier can be changed. In fact, if for any reason the barcode does not generate a New Array Identifier, enter one if you intend to send the MAGE- ML results file to Resolver.
5 To change the path for the cropped image:
Click the ellipsis (...)button.
Select the new path.
Click OK
6 Click Save Modified Image.
If the image file is from a non-Agilent microarray:
Click File > Save Modified Image to save the cropped image to a new file.
For microarrays without an identifier, the New Array Identifier field is blank. You can add an identifier. See To add or change the New Array Identifier on page 263.
Figure 81 Save Modified Image dialog box for non-Agilent microarray
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To crop an image
To add or change the New Array Identifier
Feature Extraction 12.2 User Guide
When you attempt to save a modified image, the Save Modified Image dialog box appears with a field for New Array Identifier. Agilent barcodes automatically generate this number. For non- Agilent images, this field is blank.
If the New Array Identifier is not generated, for whatever reason, the program does not transfer MAGE- ML or GEML results files to a third- party program such as Resolver.
You can change or add this identifier in the Scan Image Properties dialog box.
1 Click View > Image File Info to bring up the Scan Image Properties dialog box.
2 Click the cell next to the Identifier field, and enter the new number.
3 Click Apply, and click Close.
A message appears that says click File > Save As before the change is accepted.
4 Click OK, and click File > Save As.
5 Enter a new name or select the old name of the .tif file, and click Save.
If you were trying to save a modified image, try again. The New Array Identifier you entered appears in the field.
263
5 Changing Image Displays Loading Feature Extraction Visual Results
Loading Feature Extraction Visual Results
264
You can select to store visual results in the same local folder as the other types of result files or to have it be sent via FTP to a remote directory. In either case the visual results file is saved with the same name as the results file name with an extension of .shp.
By default, the program shows the results for outlier features only when you load the visual result file or shape file (.shp) for the image. To view all results, clear the View Outliers Only check box. It is recommended, the you first crop out a region of the image before clearing the outlier check box. For instructions, see To change the magnification (zooming) on page 279 and To change the extraction results display on page 282.
1 Open the image for which you want to see the visual result.
Figure 82 Image of Agilent 1-color, 4x44K microarray
2 Click Feature Extraction > Load Visual Result to load previously saved visual results for the currently displayed image file.
The program opens the folder containing the image file whose visual results you want to view. If you chose to save the output files for the project to a results folder separate from the image file folder, move to that folder.
3 Select the shape file, and click Open.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To view visual results for low-resolution images
Feature Extraction 12.2 User Guide
Or, you can also open the visual result file that is associated with the extraction set from the Project Explorer.
1 From Project Explorer, right- click the image.
2 Click Load Visual Results.
This option opens the image and loads the shape file associated with that extraction set.
To view visual results for low-resolution images
1 After the .shp file is loaded to the image, zoom in until you see enough of the spots to help you evaluate the results.
2 Pass cursor to feature of interest and hold for a few seconds.
A description of the feature appears.
To find a specific feature, use Ctrl-F.
Enter the Feature number of the
specific feature you want to find.
Figure 83 Visual results for low-resolution microarray image
265
5 Changing Image Displays To view visual results for low-resolution images
To view visual results for high-resolution images (2- by 3-micron or 20-bit)
266
After you first load the image, you, in fact, are viewing a preview image. You can only view the original high- fidelity image if you crop the size of the image to one- tenth or less of the original size.
Figure 84 Preview display of high-resolution image with outliers in blue
1 After the .shp file is loaded to the image, zoom in until you see enough of the spots to help you identify the region of interest.
2 Click the left mouse button at the point where you want to start forming a rectangle for cropping.
3 With the crop cursor, draw a rectangle equal to or less than one- tenth the size of the image.
The crop cursor is black if the size is within one- tenth (200% to 1000%) of the original size, but it turns red if you are drawing outside the high fidelity range.
4 Then release the mouse button.
Note the status bar at the bottom of the window. It now reads High Fidelity.
High fidelity mode is when you zoom in the image by at least 200%.
5 Pass cursor to feature of interest and hold for a few seconds.
A description of the feature appears.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To view visual results for low-resolution images
Feature Extraction 12.2 User Guide
Figure 85 Visual results of high-resolution microarray image in
High Fidelity mode, zoomed to 500%
View pixels used
This view option for low- resolution images and for high- resolution images that are in High Fidelity mode (that is, when the image is zoomed in to one- tenth of its original size).
The View Pixels Used option displays the pixels used to quantify the feature signal within the cookie area. Pixels that are determined to be outliers are shown in black.
1 If the image is a high- resolution image, complete steps 1 through 4 from To view visual results for high- resolution images (2- by 3- micron or 20- bit) on page 266.
2 Under View > Extraction Results, select View Pixels Used.
267
268
5 Changing Image Displays To view visual results for low-resolution images
Figure 86 View pixels used for high-resolution microarray image in
High Fidelity mode, zoomed to 500%
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To view visual results for low-resolution images
To save an image (original or cropped) for import to another program
Feature Extraction 12.2 User Guide
1 Click File > Save As to produce high- resolution images.
The Save As dialog box lets you save .tif files or .jpg files of the current image or cropped image.
2 Click the down arrow to the right of the Save as type: list.
3 Select either Tif Files, BMP files Png Files, or Jpeg Files.
4 Enter the name of the image.
5 Click Save.
Figure 87 Save As dialog box
269
5 Changing Image Displays To view visual results for low-resolution images
To display file information
270
You can use the File Info dialog box to view valuable information, such as scanner warning messages, scanner version, PMT voltage, dark offset information, the scan resolution, and glass thickness. For images scanned with the Agilent D Scanner, information also includes Laser set points and/or actual laser power readings for each channel (reflects compensation, if invoked), and PMT Calibration Error flag status.
Click View > Image File Info to display a tabbed dialog box that displays all available information about an image file.
Figure 88 Scan Image Properties dialog box
The only field that you can change is the Identifier field. See To add or change the New Array Identifier on page 263.
For images produced on a NimbleGen, Molecular Devices, or Innopsys
Properties dialog box.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 Changing Image Display Settings
Changing Image Display Settings
Changing the image display does
not affect the Feature Extraction
results in any way.
Feature Extraction 12.2 User Guide
When you open an image you can manipulate displayed images from the original scan file, the cropped scan file or the Feature Extraction results in the following ways:
Change image orientation (landscape/portrait)
Change the number of channels displayed
Change the display color
Change between log and linear color scale
Change the intensity scale (color display range)
Change the magnification (zooming)
Change the fit of the window to the image
Change the extraction results display
This section also shows you how to change settings for histograms and line graphs, the default image analysis parameters and printing.
To change image orientation (landscape/portrait)
Due to the size of 2um or 3um scan
images, the flip and rotate
operation can take several
minutes.
When you first load a microarray image into the Feature Extraction Tool, the image appears in the orientation produced by the scan, either landscape or portrait. If the gene list that you want to use to position a grid and find spots has a different orientation than the image, change the image orientation to match that of the gene list.
Flip and rotate of 2um, 3um, and
20-bit scans is supported in v10.7
and above.
1 Click Tools > Flip Upper Left to Lower Right (Landscape/Portrait) from the top menu.
If the original orientation was landscape, the program takes the upper left corner and moves it to the lower right corner to create a portrait orientation. The same is true if you go from a portrait to landscape orientation.
2 Click File > Save modified image.
If you do not save the image in its new orientation, you could run into problems with Feature Extraction later.
271
5 Changing Image Displays To change the channels that display
To change the channels that display
272
The default setting for displaying the channels is to display one window with the two color channels combined.
To change the channels for a single window:
1 On the image viewing toolbar, click the down arrow for the Channel Display menu.
The default selection is All Channels.
2 Select Red Channel or Green Channel.
The window display changes to the channel selected.
You can also change the number and color of the channels displayed when you set default image settings.
To display separate channel images:
Displaying the images of separate color channels side by side can be useful for comparing background anomalies on the microarray.
1 Click Window > New Filename View Window.
Filename represents the path and name of the image that you loaded. The Select Image dialog box appears.
2 Click Red or Green.
3 Click OK.
If you repeat steps 1 through 3, all three windowsthe red, the green, and the combined colordisplay at the same time.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change the channels that display
To change the display color
Color Mode choices are False
Color, Gray Scale, Reverse Gray
Scale and False Color w/White BG.
Feature Extraction 12.2 User Guide
From the image viewing top menu, click Color > color mode.
Figure 89 Color menu
The color type that you select depends on the channel image display. For combined channels, there are two choices of color, and for single channels there are three choices.
Combined channel display
The default setting for the color on the TIFF image is called False Color. You also have the option of selecting False Color with a white background.
Click Color > False Color w/ White BG.
To return to False Color, click Color > False Color.
False Color
False Color w/ White BG
This selection maps intensities the same as the False Color palette, except that a pixel with zero values in both red and green is turned white instead of black. This feature is useful when printing because it saves ink. To produce the best possible image, manually adjust the minimum data range.
273
5 Changing Image Displays To change the channels that display
274
Single channel display
The default setting for the color on the single- channel display is also False Color.
Click either Color > Gray Scale or Color > Reverse Gray Scale.
Gray Scale
Reverse Gray
To return to False Color, click Color > False Color.
Examples of images with different color selections
The following images are intentionally not aligned.
Figure 90 False Color for 2-color image
Figure 91 False Color with white BG for 2-color image
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change the channels that display
Feature Extraction 12.2 User Guide
Figure 92 False Color for 1-color image
Figure 93 Gray Scale
Figure 94 Reverse Gray Scale
275
5 Changing Image Displays To change the intensity scale (color display range CDR)
To change between log and linear color scale
276
Click Color > Use Log Color Scale to toggle the color scale from linear to log10 (intensity).
To return to the linear scale, click Color > Use Log Color Scale again to remove the check mark.
With the log scale, you can see low intensity portions of the microarray at the expense of detail at high intensities.
You can also use the Log/Linear button on the toolbar to toggle the color scale.
To change the intensity scale (color display range CDR)
Changing the color display range affects how the image intensity is scaled. Modifying the color display range can make it easier to visualize your microarray image but does not affect the raw intensity.
You can change the color display range in one of two ways:
Using the color scaling toolbar.
Using the Set Color Display Range dialog box.
Use the color scaling toolbar
The easiest way to determine the best range is to use the Auto Scalefunction on the toolbar.
Figure 95 Color scaling segment of the image viewing toolbar
CDR Low Mode Icon CDR High Mode Icon
Feat
Changing Image Displays 5 To change the intensity scale (color display range CDR)
Feature Extraction 12.2 User Guide
1 Click the down arrow next to the color scaling menu, and click Auto Color Scaling if necessary.
2 Click the CDR Low Mode icon, and move the lower bar cursor to the minimum percentage for the display range.
3 Click the CDR High Mode icon, and move the higher bar cursor to the maximum percentage for the display range.
Notice the change in color of the image display as you increase or decrease the blue bar size.
4 To fine- tune the color display ranges manually, click either the Red Channel or the Green Channel from the color scaling menu.
5 Click the CDR Low Mode icon, and move the lower bar cursor to the minimum pixel number for the display range.
6 Click the CDR High Mode icon, and move the higher bar cursor to the maximum pixel number for the display range.
277
278
5 Changing Image Displays To change the intensity scale (color display range CDR)
Use the Set Image Color Display Ranges dialog box
Alternatively, you can set the color display range through the Set Image Color Display Ranges dialog box.
1 Click Color > Set Color Display Range, or click Set Color Range on the toolbar. .
Figure 96 Set Image Color Display Ranges dialog box
2 Mark the Auto Scale Image check box to set a low and a high percentage cutoff value.
Setting 0 for a low value and 100 for a high value compresses the entire range of possible values into the 8- bit displayable range. If this setting produces a black image, try 1 and 99 (the default settings). Generally, this setting gives you an acceptable data range, even for very poor images.
3 To fine- tune the color display ranges manually, clear the Auto Scale Image check box.
4 Change the Minimum and Maximum Red and Green values.
The data range determines how the 16- bit image data is mapped to the 8- bit data that is displayed on a computer screen. Generally, scanned images do not cover the entire range of possible values: 0 to 65535. Therefore, if you try to display the entire range, all you can see is black.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change the magnification (zooming)
To change the magnification (zooming)
To set the default % zoom settings
for the loaded image and the grid
mode images, see To change
default image viewing settings on
page 287 of this guide.
Feature Extraction 12.2 User Guide
Magnification can only be set to one of the % zoom levels shown in Figure 97. Numbers greater than 100% map one scan pixel to multiple pixels on the screen. For example, 600% maps one scan pixel to a 6 x 6 square of identical pixels. A zoom value of 100% maps one scan pixel to one pixel. Values less than 100% display the mean value for several scan pixels to one pixel. For example, 50% averages four scan pixels, a 2 x 2 matrix, and displays the average as one pixel.
You can control zooming in one of three ways:
Dragging the cursor to make a rectangle
Using buttons on the tool bar and on the mouse
Using the View menu
Drag the cursor to make a rectangle
You can only zoom the image using this method if cropping is turned off.
1 Clear the Cropping Mode On/Off button to turn cropping off.
2 Position the cursor at the top left corner of the image section that you want to zoom, and hold the left mouse button down.
3 Drag the cursor to the opposite corner of the section you want to zoom.
4 Release the left mouse button to display the selected area.
Cropping mode OFF
279
280
5 Changing Image Displays To change the magnification (zooming)
Use the Tool Bar and Mouse Buttons
Three buttons on the tool bar and the right and left mouse buttons control zooming. You can use these buttons with Crop Mode On or Off.
Use the Zoom menu
You can use the Zoom menu with Crop Mode On or Off.
Click View > Zoom to jump to any zoom level directly.
For low- resolution images, if the image is uncropped, the zoom levels range from 10% to 1000%.
For high- resolution images, if the viewed image is uncropped, the zoom levels range from 10% to 100%. To activate zoom levels from 200% to 1000% (High Fidelity mode), crop the image.
The crop cursor is black if the size is within one- tenth (200% to 1000%) of the original image size, but it turns red if you are drawing outside the High Fidelity range.
Figure 97 Zoom menu for low-resolution image
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change the window size to fit the image
To change the window size to fit the image
Feature Extraction 12.2 User Guide
If you have to scroll to see the entire image, use this option to increase the size of the window to fit the image. If the window is too large for the image and contains white space around the perimeter of the image, use this option to decrease the size of the window to fit the image.
Click View > Entire Image to resize the window.
281
5 Changing Image Displays To change the window size to fit the image
To change the extraction results display
282
Use this option to control the portions of the feature extraction results that are displayed.
Click View > Extraction Results.
Figure 98 Extraction Results Default views selected
View Results
Shows all visual results when this option is selected and View Outliers Only is not selected.
Shows outlier results only when both this option and View Outlier Only are selected.
Hide all visual results when this option is not selected, even if View Outliers Only is selected.
View Outliers Only
Hide Outer Local BG Ring
Shows or hides the local background ring. See To select a spot statistics method to define features on page 209 of this guide for a definition of the local background.
Use Simple Colors
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change and save histograms and line graphs
View Pixels Used
Feature Extraction 12.2 User Guide
Shows or hide pixels within the cookie that are used to quantify the feature mean or median intensity.
For more details, see View pixels used on page 267.
To change and save histograms and line graphs
For high- resolution images, histograms and line graphs use the original data, not the data from a low- resolution preview image.
Set up a histogram graph
Histograms can be very useful to visualize your results. They can show you if you have an anomalous distribution of intensities across the entire slide or across rectangular or elliptical subregions of the image.
Entire viewing area
1 Click View > Entire Histogram from the main menu bar.
If the selected file has more than one color, a Channel Selection dialog box appears.
2 Select the desired channel, or select both.
Rectangular subregion
1 Make sure that cropping is OFF.
2 Click the right mouse button to position your cursor in one corner of the region to capture.
3 Hold down the right button and drag the pointer to the opposite corner, then release.
If the selected file has more than one color, a Channel Selection dialog box appears.
4 Select the desired channel, or select both.
283
284
5 Changing Image Displays To change and save histograms and line graphs
Elliptical subregion
1 Click Tools > Preferences > Graph Viewing.
2 Click the cell next to Select Elliptical Area for Histogram, and click the Up arrow.
3 Select True, and click OK.
False is the default setting and results in a rectangular subregion.
Figure 99 Histogram Preference Settings dialog box
4 Make sure that cropping is OFF.
5 Click.
Or, click View > Select Elliptical Area.
6 Click the right mouse button to position your cursor in one corner of the region to capture.
7 Hold down the right button and drag the pointer to the opposite corner, then release.
If the selected file has more than one color, a Channel Selection dialog box appears.
8 Select the desired channel, or select both.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change and save histograms and line graphs
Feature Extraction 12.2 User Guide
Change the bin size for histograms
1 To adjust the bin size after creating the initial graph, click Histogram > Change Bins.
2 Change the bin size or the maximum number of bins.
3 Click OK.
The maximum bin size is 65535; the minimum is 1.
The maximum number of bins is 32768; the minimum is 2.
Change the default bin size for histograms
1 Click Tools > Preferences.
2 Click the Graph Viewing folder.
3 Change the bin size or the maximum number of bins.
4 Click OK.
285
286
5 Changing Image Displays To change and save histograms and line graphs
Set up a line graph
Line graphs show a vertical or horizontal slice through the image. They can be useful for spotting extreme differences between the red and green signal intensities across the microarray. Line graphs are also useful for displaying average intensity over all rows and columns, minimizing the effect of noise.
To create a line graph more quickly,
double-click a selected row within
the image with the left mouse
button, or double-click a selected
column with the right mouse
button.
1 Click View > Select Row or Column, or click or .
The Select Row/Col dialog box appears.
Figure 100 Select Row/Col dialog box
2 Click Row Number for showing signals in a row, or click Column Number for showing signals in a column.
3 Enter the number of the row or column.
4 Mark the All Rows or Columns check box to average the rows or columns in the view depending on the Row Number or Column Number selection.
5 Click OK.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change default image viewing settings
Feature Extraction 12.2 User Guide
Figure 101 Line graph example
Use the graph toolbar
You save graphs produced from the image files to third- party programs, such as word processors and spreadsheets, with the buttons in the Graph Toolbar. You can also print and modify the graph display with this toolbar.
Click the selected button in the Graph Toolbar at the top of the histogram or line graph.
To change default image viewing settings
You can also change the default settings that appear when you open a new file. To activate the settings, exit the image file and either reopen the same file or open a new one.
287
5 Changing Image Displays To change default image viewing settings
See To change the display
color on page 273, To change
between log and linear color
scale on page 276, and To
change the intensity scale (color
display range CDR) on page 276.
288
1 Click Tools > Preferences.
2 Open the Image Viewing folder, if necessary.
3 Change the default selections if necessary.
4 Click OK.
5 Close the image.
6 Reopen a file.
The new default settings now take effect.
Figure 102 Image View Default Options dialog box
Color Palette
Use log color scale
If set to True, the program uses the log10 intensity scale for the color. If False, it uses a linear scale.
AutoScale
Start with crop mode
Select True if you want to be able to crop the image with mouse movement right after you load the image. Select False if you want the mouse to activate the zoom function.
Ratio of window size to feature cross for found
feature
When you attempt to find a feature in the visual result image, you can change the ratio of the resulting window size for the image to the size of the feature cross for this found feature.
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To change the print settings
Feature Extraction 12.2 User Guide
Change other default settings in the Preferences dialog box
See the tasks in the relevant chapters for changing the default settings for the Grid Mode (Chapter 2), for Graphs (histogram bin settings in this chapter) and for Projects (Chapter 1).
To change the print settings
The Feature Extraction application supports printing to both color and black and white printers.
Change page settings
Click File > Page Setup to set the following printer paper parameters:
Paper size
Paper source
Paper orientation
Margins
Printer
Change print settings
1 Click File > Print Setup or File > Print to bring up the standard Windows Print Setup dialog box.
Within this dialog, you can change the following parameters:
Paper size
Paper source
Paper orientation
Margins
Printer
2 Click OK on the Print Setup dialog box to save changes to the printer setup without printing, or
289
290
5 Changing Image Displays To change the print settings
Click Print to save changes to the printer setup and to print.
Preview a file before printing
Click File > Print Preview.
Printing single color images
If you are printing a single color image that is displayed using the false color palette to a black and white printer, the Image Analysis Tool of Feature Extraction automatically converts the image to gray scale. This conversion is necessary because the gray scale intensities that a black and white printer automatically generates for a false color image are incorrect.
Examples
Figure 103 False Color image
Figure 104 Incorrect image generated by blank and white printer
Figure 105 Correct gray scale generated by the Image Analysis Tool
Feature Extraction 12.2 User Guide
Changing Image Displays 5 Hot key actions for image displays
Feature Extraction 12.2 User Guide
Printing 2-color images
Printing 2- color images on a black and white printer does not require the Image Analysis Tool to convert the image to gray scale, but you cannot distinguish between the colors. See the following example.
These images are intentionally not
aligned.
Figure 106 2-color image
Figure 107 2-color image printed on a black and white printer
Images always print at maximum resolution. The zoom level does not affect printing. This resolution setting makes printing the image slow but also generates the highest- quality image possible.
Hot key actions for image displays
Table 11 Hot key actions for image displays
If you want to: Press this hot key:
Turn Crop Mode On/Off Alt-C
Call up the manual crop dialog box Ctrl-M
Call up the File Info dialog box Ctrl-I
Change image orientation
(landscape/portrait)
Shift-U
Zoom in Ctrl-double-click left mouse
291
292
5 Changing Image Displays Hot key actions for image displays
Find feature when viewing extraction
results
Ctrl-F
Open new window
Does not work for high-resolution
scans (2um, 3um, or 20-bit)
Ctrl-N
Open file Ctrl-O
Print TIFF image Ctrl-P
Save the image as a different file Ctrl-S
Table 11 Hot key actions for image displays
If you want to: Press this hot key:
Feature Extraction 12.2 User Guide
Changing Image Displays 5 Changing Feature Extraction Window Displays
Changing Feature Extraction Window Displays
Feature Extraction 12.2 User Guide
When you start the Feature Extraction program, you see the default window layout that Agilent configured before shipment. This window layout always appears unless you move windows and panes and change the default settings in the Preferences dialog box.
Figure 108 FE Standard Project Configuration window default
Project Work Window Protocol Browser
Pane
Grid Template Browser Pane
Project Explorer Pane
Restore Window Icon
QC Metric Set Browser Pane
293
294
5 Changing Image Displays Changing Feature Extraction Window Displays
The following figure shows another layout of the FE Standard Project Configuration window. This section shows you how to change from one layout to a new layout and to save the new layout so that it appears when the program starts.
Figure 109 FE Standard Project Configuration window new layout
Feature Extraction 12.2 User Guide
Changing Image Displays 5 Changing Feature Extraction Window Displays
To move windows and panes
Feature Extraction 12.2 User Guide
Windows can be separated from window frames and blended into frames (maximized).
Click the Restore Window icon for the Project Work Window. See Figure 108.
The Project Window becomes separated from the main window frame, exposing the Workspace.
Panes can be unattached from one edge of a window frame and docked to another edge of the frame.
1 Click on the title of the Grid Template Browser Pane and drag it from the right edge of the window frame.
2 Move the Browser across the Workspace until the cursor hits the left edge of the frame, then release the mouse button.
The Grid Template Browser is now docked to the left edge of the frame. You can dock a pane to any edge of the frame in this way.
3 Repeat step 1 for the FE Protocol Browser Pane.
4 Move the Browser across the Workspace and place the cursor on the title of the Grid Template Browser.
Now both panes are overlapping one another with tabs at the bottom of the panes. To separate the two panes once more, click on a tab and move it to a new position.
If you had moved the FE Protocol Browser Pane to the left edge of the frame, the two Browsers would be next to each other.
5 Click on the title of the Project Explorer Pane and drag it across the Project window until the cursor hits the right edge of the window frame, then release the mouse button.
Your screen now looks like Figure 109.
Two other actions can be taken with panes:
Close panes. Click on the X in the upper right corner.
Autohide panes when docked. Click on the pin icon in the upper right corner of the pane.
295
296
5 Changing Image Displays To select the layout to use when program starts
The pane now appears as a bar button on the edge to which it was docked.
When you click the resulting bar button, the pane floats from the edge so that you can view its contents. The pin position is now horizontal instead of vertical. If you click the pin now, the pin becomes vertical and the pane redocks itself to the edge of the frame.
To select the layout to use when program starts
You set the default layout for the Feature Extraction display in the Preferences dialog box.
1 Click Tools > Preferences from the top menu.
2 Change the following settings to produce the layout you want upon program startup.
If you want to return the windows
and panes to their original
positions at any time, set Frame
Windows Using last saved
Position and Docking Pane Using
last saved Layout to FALSE, close
Feature Extraction and reopen the
program.
Figure 110 Preference Settings dialog box
Feature Extraction 12.2 User Guide
Changing Image Displays 5 To select the layout to use when program starts
Frame Windows Using last saved Position
Feature Extraction 12.2 User Guide
If True, the window positions you have changed in the last session appear upon startup. If False, the factory default window position appears. To have Figure 109 appear upon startup, keep the setting as True.
Docking Pane Using last saved Layout
If True, the docking positions from the last session appear upon startup. If False, the factory default docking positions appear. To have Figure 109 appear upon startup, keep the setting as True.
Initially Loading Project
If set to New Project, FE Project1 always appears upon startup. If Last Project, the project on the screen before exiting appears at startup. If None, the Workspace is blank upon startup. Keep New Project to reproduce Figure 109.
297
298
5 Changing Image Displays To select the layout to use when program starts
Feature Extraction 12.2 User Guide
Index
Index
A access, 119
additive error, 229
Agilent eArray, update grid info, 85
algorithm
purpose, 29
B background correction
adjust background globally, 227
background subtraction method, 222
multiplicative detrending, 221, 222
no background subtraction, 222
spatial detrending, 225
surface fit calculation, 225
background subtraction methods
global, 222
local, 222
barcode id, add to tiff header, 263
C calculate spot size and centroids, 163
channels
changing display, 272
colors
changing log/linear scale, 271, 276
changing the display, 272
examples of selections, 273
Cookie Cutter method, 209
correct dye biases
normalization correction method, 241
positive and negative controls, 240
select probes, 239
select signal characteristics to determine rank consistency, 241
D default settings
grid mode, 176
Feature Extraction 12.2 User Guide
default settings, changing for Image Viewing, 287
definitions
dye bias, 237
feature, 205
log ratio, 243
log ratio error, 243
negative controls, 224
p-value, 243
spot, 205
differential expression
p-value, 254
display settings
changing colors, 287
changing defaults, 287
dye bias
definition, 237
removing differences, 237
dye normalization
algorithms, 241
methods of curve fitting, 241
selecting set of features, 239
what algorithm does, 241
DyeNorm gene list
change default, 93
create, 93
create external, 97
edit, 97
Editor, 94, 100
export, 97
external file, 54
import, 94, 100
save, 97
E error models
combination, 243
propagated error, 243
Universal, 243
error propagation
correct bkgd and signal biases, 234
find and measure spot, 212
extraction results display, changing, 282
extraction results, viewing, 281
extraction sets
access grid mode, 134
access grid mode from, 85
create new project if run fails, 133
create project if run fails, 114
definition, 71
view number in project, 49
F Feature Extraction
1-color run, 113
monitor runs, 67
number set to run simultaneously, 67
start, 113
stop, 113
terminate, 47
turn on/off monitoring, 113
Feature Extraction algorithms, 182
features
define with Cookie method, 209
test to assess if positive and significant, 234
file information
displaying, 270
FTP results, 52
G gene list type
gal format file, 144
no gene list, 144, 148
tab text file, 144, 145
geometry information, 148
grid
hot key actions, 177
position main grid, 151
position spot centroids, 167
refit, 174
299
Index
set up initial, 130
shortcuts for actions, 177
zoom in or out, 151
Grid Adjustment pane
move main grid, 161
position spot centroids, 172
position subgrids, 163
grid editing
change defaults, 176
grid file
add to extraction set, 84
change in extraction set, 84
definition, 72
use if available, 54
grid mode
access from extraction set, 134
access from image file, 136
grid mode, change default settings, 176
grid placement
allow some distortion, 196
failures, 118
place and rotate only, 196
summary report warnings, 118
grid template
add to database, 84
add to extraction set, 83
change default protocol, 91
change in extraction set, 83
definition, 72
properties, 90
update via Agilent eArray, 85
gridding
if fails for Agilent image, 132
H histograms
changing bin size, 285
purpose, 287
selecting elliptical area, 287
selecting elliptical or rectangular region, 283
setting up, 283, 287
viewing, 287
hot key actions
300
grid mode, 177
image analysis, 177
I ignore spots, 171
image
cropping, 259
hot key actions, 291
printing, 289
saving cropped, 261, 262
viewing entire image, 281
viewing extraction results, 281
image display
changing, 271
image orientation
changing, 270
image, changing window size to fit, 281
initial grid, 130
intensity scale, changing, 276
IsPosAndSignif t-test, 234
L landscape, change to portrait, 142
line graphs, creating, 286
linear scale, changing to log, 276
local background
define with auto-estimate of local radius, 211
definition, 211
log ratios
error models, 229
surrogates, 230
log scale, changing to linear, 276
M magnification levels, 279
main grid
move position, 156
move with Grid Adjustment pane, 161
rotate position, 156
microarray information, process flow, 18
multiplicative error, 229
N negative controls
definition, 224
New Array Identifier, add or change, 263
O orientation, portrait or landscape, 142
outliers
automatically compute polynomial terms, 216
features and local background, 214
interquartile range method, 215
non-uniformity, 215
parameters for polynomial on, 216
population, 215
output files
GEML, 54
grid, 54
JPEG, 54
MAGE, 54
QC Report, 54
text result, 54
visual results, 54
P pixel outliers
method to reject, 211
pixel saturation, 212
portrait, change to landscape, 142
position
main grid, 151
spot centroids, 167
Preferences dialog box
changing other default settings, 289
grid mode, 176
print settings
changing, 289
printing images, 289
process flow
microarray information, 18
project
change default properties, 57
Feature Extraction 12.2 User Guide
Index
change specific properties, 45
create new if run fails, 114, 133
default protocol, 53
folders, 59
load visual results, 121
on-time, 35
open, 43
save, 44
standard, 35
summary report, 117
terminate, 47
project properties
differences between standard and on-time, 47
protocol, 182
1-color, 91
1-color gene expression, 187
2-color Agilent gene expression, 182
2-color gene expression, 187
add new protocol, 192
add to extraction set, 105
Axon gene expression and CGH, 187
CGH, 182, 187
change in extraction set, 105
change in grid template, 91
change settings, 105
change steps to run, 187, 192
create from existing protocol, 184
definition, 72
derived from, 186
export, 105
hidden settings, 184
import, 105
location for default, 53
new from template, 184
new, from template, 183
open, 185
project default, 53
properties, view or change, 186
removable from database, 187
remove from database, 105
save as permanent read-only, 192
save to new name, 192
templates from Agilent, 182
type, 187
Feature Extraction 12.2 User Guide
version, 186
protocol steps, 182
brief description, 188
calculate metrics, 252
compute ratios and errors, 243
correct dye biases, 237
Feature Extraction algorithms, 182
find and measure spots, 204
flag outliers, 213
generate results, 255
place grid, 193
p-values
surrogates, 230
Q QC Report
1-color, 51, 63
2-color, 51, 63
access from Project Run Summary, 118, 119
CGH, 51, 63
open, 119
print, 121
type, 51, 63
view, 119
QC Report metrics
p-value, 254
replicate statistics, 253
spikein targets, 253
R radius method for background
measurement, 211
rank consistency
use in dye normalization, 241
refit a grid, 174
replicate statistics
minimum population, 253
results
FTP to server, 64
jpeg compression factor, 256
loading visual, 264
overwrite, 54
save to folder, 51, 63
text output as single file, 255
text output as three files, 255
text result file, 115
Rosetta Resolver
ACL settings, 50, 62
data access settings, 50, 62
GEML transfer, 49, 61
MAGE transfer, 49, 61
send FTP results, 52
S saturation
definition, 212
saturation, pixels, 212
save grid file, 165
saving images
cropped, 261, 262
scan image properties, 270
shortcuts, find spots, 177
significance tests
2-sided t-test, 234
well above SD, 234
software features, 29
spike-in targets, 252
spot analysis
CookieCutter method, 209
WholeSpot method, 209
spot centroids
calculate, 163
position, 167
position with Grid Adjustment pane, 172
Spot Navigator, 169
spot size
calculate, 163
spots
calculate size, 204, 208
Cookie statistics method, 209
find and measure, 204
ignore, 171
statistics method to define features, 209
undo or redo position, 171
301
Index
WholeSpot statistics method, 209
subgrid
position with Grid Adjustment pane, 163
skew, 163
surrogates
effect of use, 246
T t-test
IsPosAndSignif, 234
two-sided, 234
U Universal Error Model
additive error, 243
multiplicative error, 243
V visual results
load from a project, 121
view, 121
W WholeSpot method, 209
window layout
default, 293
Grid Template Browser Pane, 293
Project Explorer Pane, 293
Project Work Window, 293
Protocol Browser Pane, 293
restore, 297
set preferences, 297
when software starts, 296
Workspace, 293
windows and panes
moving, 295
Z zoom
grid mode, 151
302
zooming
tool and mouse buttons, 279
Feature Extraction 12.2 User Guide
www.agilent.com
In This Book
The User Guide presents instructions for using Agilent Feature Extraction 12.2.
This guide provides instructions for:
Setting up and running Feature Extraction on batches of image files.
Creating grid files to assign to image files for automatic extraction.
Changing the parameter values for each step in a protocol that can be assigned to an image file for automatic extraction.
Working with image displays.
Related manuals for Agilent Feature Extraction 12.2 Software User Guide
Manualsnet FAQs
If you want to find out how the Feature Extraction 12.2 Agilent works, you can view and download the Agilent Feature Extraction 12.2 Software User Guide on the Manualsnet website.
Yes, we have the User Guide for Agilent Feature Extraction 12.2 as well as other Agilent manuals. All you need to do is to use our search bar and find the user manual that you are looking for.
The User Guide should include all the details that are needed to use a Agilent Feature Extraction 12.2. Full manuals and user guide PDFs can be downloaded from Manualsnet.com.
The best way to navigate the Agilent Feature Extraction 12.2 Software User Guide is by checking the Table of Contents at the top of the page where available. This allows you to navigate a manual by jumping to the section you are looking for.
This Agilent Feature Extraction 12.2 Software User Guide consists of sections like Table of Contents, to name a few. For easier navigation, use the Table of Contents in the upper left corner.
You can download Agilent Feature Extraction 12.2 Software User Guide free of charge simply by clicking the “download” button in the upper right corner of any manuals page. This feature allows you to download any manual in a couple of seconds and is generally in PDF format. You can also save a manual for later by adding it to your saved documents in the user profile.
To be able to print Agilent Feature Extraction 12.2 Software User Guide, simply download the document to your computer. Once downloaded, open the PDF file and print the Agilent Feature Extraction 12.2 Software User Guide as you would any other document. This can usually be achieved by clicking on “File” and then “Print” from the menu bar.