Using PhenoGen to Analyze Your Microarray Data:

1
Using PhenoGen to Analyze Your Microarray Data

• A Case Study In Going From Microarray Data to
  Candidate Genes

You will use an analysis of 21 BXD RI strains
mRNA microarrays along with the morphine
analgesia phenotype data from Belknap (1995) to
identify candidate genes for morphine-induced
analgesia.
2
Overall Analysis Scheme

• Correlate Phenotype and Gene Expression Data
• Pre-filter the Microarray Data
• Perform an Affy Control Genes Filter
• Perform a MAS5 Absolute Call Filter
• Perform a Microarray Data Analysis to Estimate
  the Correlation of Each Transcript With the
  Phenotype
• Create Candidate Gene List
• Further Filter Candidate Gene List
• By Phenotypic QTL (pQTLs) regions and eQTLS

3
Performing a Microarray Data Analysis
Click on the Analyze Microarray Data Tab
4
Performing a Microarray Data Analysis (cont.)
Click on the Analyze microarray data Link
Click here
5
Select A Dataset To Analyze
Click on the Public BXD RI Mice dataset
This is the dataset that you will be analyzing
Click on it
6
Select A Normalized Dataset
Click on View to see the dataset attributes
7
View Dataset Attributes
Click to expand for more details
Click the minus icon to retract this group list
8
Select A Normalized Dataset
Click on the rma and probe mask-normalized
version 6
9
Correlate Phenotype and Gene Expression Data
Select the Correlation type of analysis
Take note of this schematic to show you where you
are in the process
10
Entering Phenotype Data for a Correlation Analysis
Click on the Enter New Phenotype Data button
11
Enter New Phenotype Data
Select the Enter New Phenotype Data option
12
Enter New Phenotype Data
Enter the phenotype data values for the strains
indicated as you see below, then click the Save
New Values button.
Click here when finished entering strain means
13
Save New Phenotype Data
Note that means for only 21 strains were loaded
Close the Message window
14
Select Phenotype Values
Select the morphine analgesia row from the
Phenotype Values table
15
Filtering Probes Affy Control Genes Filter
Select the Affy Control Genes Filtering method
Then Click here to Run the Filter
16
Filtering Probes MAS5 Absolute Call Filter
Select the MAS5 Absolute Call Filtering method
Select Probes present in at least 50 of the
samples
Note of probes remaining after the Affy Control
Gene Filter finished running
17
Filtering Probes MAS5 Absolute Call Filter
Click Next to go to the statistical analysis
page
Note that 26,238 additional probes were
eliminated with the MAS5 Absolute Call Filter
18
Running a Correlation Analysis
Select the Pearson method of Correlation
analysis
19
Multiple Testing Adjustment
Select No Adjustment and alpha 0.05
20
Correlation Analysis Results
Note 866 Statistically Significant Genes at
alpha 0.05
21
Accessing Your Gene List
Note Click on the Research Genes or
Investigate QTL Regions tab to access your gene
list(s)
22
Accessing Your Gene List
Select a gene list for further investigation
Click here
23
Accessing Your Gene List
Click View for the BXDRI_AffyMAS5alphaFilterCor
rWmorphineAnalgesia gene list
24
Gene List Details
Click the and icons to expand and retract the
details
25
Your Gene List
Click on the Location tab to map your gene list
26
Your Gene List Visually Displayed on a
Chromosomal Map
Check this box then select Advanced Settings
27
Creating a New pQTL Region
Click here to create a new pQTL region
28
Creating a pQTL Region
Fill in the information for your new pQTL region,
then save it
29
Selecting Your pQTL
Click on your pQTL region to select it
30
Viewing Your Gene List
Select a criterion for viewing your gene list
31
pQTL-Filtering Your Gene List
Note The genes shown on the map are either
physically located in the pQTL (bracketed)
regions or the area that controls their
transcription (or its 95 confidence intervals)
overlaps the pQTL regions
32
Saving Your pQTL-Filtered Gene List
Name and describe your gene list, then save it
33
Viewing Your pQTL-Filtered Gene List
View your gene list in table form
Click here
34
Viewing Your pQTL-Filtered Gene List
Affy gene identifiers, gene symbols and best
markers are among the information given in the
gene list table.
35
eQTL-Filtered Candidate Gene List
Candidate genes are shown on the chromosomal map
36
Saving Your eQTL-Filtered Candidate Gene List
Click on the Save gene list link
37
APPENDIX
The following exercises are shown as examples in
this section

• Gathering arrays to assemble a dataset
• Running a Quality Control Analysis
• Grouping and Normalizing Array Data

38
Create a New Dataset
Click on the Analyze Microarray Data Tab
39
Create A New Dataset (cont.)
Click on the Create a dataset link
Click here
40
Create A New Dataset Retrieve Arrays
Type BXD into the Hybridization Name contains
criterion box
Then click on the Get Arrays button
41
Create A New DatasetDisplay Selected Arrays
Note 168 Arrays matched the selection query
Click the Display dropdown menu to view more
arrays per page
42
Create A New DatasetRetrieve Arrays
Check boxes for the arrays you want to retrieve
Then click to Add Selected Arrays to Dataset
43
View and Finalize A New Dataset
Click the View/Finalize Dataset icon
44
View A New Dataset
Name and Finalize your dataset
45
Finalize A New Dataset
Your dataset is now finalized
46
New Dataset Status
Click Run to Run your QC analysis
47
Run a Quality Control Check
Click Yes to generate images and Run the QC
analysis
Take note of the analysis steps
48
Quality Control Check
Note the status of your dataset now
49
Quality Control Check Accessing Your Dataset
Click here to access your dataset
50
Quality Control Check Status
Note the status of your datasets QC process
51
Quality Control Check Array Compatibility
Note Strain is the only flagged attribute with
differences
52
Quality Control Check Within Array Checks
Scale value axis
53
Quality Control Check Model-Based Checks - RLE
54
Quality Control Check Model-Based Checks
Normalized Unscaled Standard Errors
55
Quality Control Check Model-Based Checks
Pseudo Images
56
Quality Control Check MA Plots
57
Quality Control CheckApprove Quality Control
Results
Click to Approve Quality Control Results
58
Grouping Datasets
Select Strain as your mode of grouping
59
Grouping Datasets
Note the Strain names and categorization are
entered automatically
Enter a name for this grouping
Then click here
60
Normalizing Your Array Data
Click to select the grouping that you just
created, then select the mas5 Normalization method
61
Normalizing Your Array Data
Enter a version name
Click here to start the Normalization process
62
Normalizing Your Array Data
Note the status of your dataset