Affymetrix Facility Documentation
Affy Facility Docs Home
Running Chips
Sample Submission
Download Data
Annotate Samples
Get Affy Intensity
Analize Data
Analysis Pipeline
Affy Help Pages
Affy Core Help
SBEAMS Help Pages
Analysis Help Pages
Tutorials and Presentations
Other Resources
Affymetrix Home Page

Affy Pipeline Analysis methods

Currently the affy analysis pipeline supports the following methods

SAM is currently the default method for array analysis. The method was developed by Tusher et al. Below are links to the Tusher Paper and some external links that help explain what SAM does.

1)Learn how to start running SAM
2)Power Point presentation from the SAM method authors
3)Original SAM Paper
4)Bioconductor's implementation of SAM "siggens"
MEV (MultiExperiment Viewer) is a stand alone, full featured array analysis package developed at TIGR which is freely distributed. The program can be easily launched from the analysis pipeline once there is some normalized data to analyze. Currently data loaded into MEV cannot be loaded back into SBEAM for additional analysis

1)View a tutorial by Bruz M. on how to start MEV from the SBEAM Affy analysis pipeline
2)Link to TIGR MEV Web Site
Make ratios simply combines the replicates within a sample sample group then makes a ratio to it's reference sample group. This method is only intended to be used on data sets that do not have replicates. After the ratios are done being made the data can be loaded back into GetExpression

1)Running the page is very similar to Running SAM. Just click on the "Make_ratios" button instead of SAM when choosing the analysis method. Click the RUN SAM link above to see how to get things started.
multtest includes several tests for differences in means (t-test, F-test, etc.) as well as a number of procedures for controlling the error rate for many simultaneous hypotheses. Such control is very important in the context of microarray experiments. Additionally, multtest provides the ability to make specific hypotheses about groups of genes within a microarray experiment and can select genes that meet certain testing criteria. Finally, muttest produces three diagnostic plots, an MA plot, a quantile-quantile plot, and a multiple testing procedure selectivity plot. After the data is done processing the differentially expressed data can be loaded into GetExpression

1)Resampling-based multiple testing for microarray data analysis.
2)Multiple hypothesis testing in microarray experiments.