AltAnalyze question

[Paula Vaz]

Hello,
I am a PhD student at the University of Sydney with a question about the output from the Altanalyze program. I am working with a set of Affymetrix exon array data to analyze changes when my gene of interest was knocked down. When we analysed the data using an in-house program, we found a set of 175 genes in which one probe set was significantly down regulated by at least ten fold though the gene expression did not change very much- and these probe sets were on or near the polyadenylation signal at the end of the 3’ UTR of the gene. When we analysed our data with AltAnalyze, these genes were not flagged as being alternatively spliced. Could you please tell me why?
I’m a biologist and not that tech savvy, so your assistance would be appreciated.
Kind regards,
Paula Vaz
--
Paula Vaz
PhD candidate
Mackay Lab
School of Molecular Bioscience,
Building G08, University of Sydney
NSW 2006, Australia

ph: +612 90367890
fax: +612 93514726

pvaz...@uni.sydney.edu.au

[Nathan Salomonis]

Hi Paul,

First off, AltAnalyze seems to have a fairly high true positive prediction rate, based on ours and others analyses:
http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1000553
http://www.altanalyze.org/help.htm#evaluation
http://code.google.com/p/altanalyze/wiki/CitingArticles

Can you tell me a little more about the method used to assess alternative exon expression used in house and other filters? How many total genes were predicted from AltAnalyze as being alternatively expressed? What parameters in AltAnalyze did you use (e.g., splicing-index versus FIRMA, constitutive versus core probesets for gene expression, expression/DABG cutoffs).

You can always change the filtering parameters of AltAnalyze to determine if it is too strict. For example, change the expression cutoff from 70 to 1. See:
http://www.altanalyze.org/help_main.htm
http://code.google.com/p/altanalyze/wiki/Constitutive_vs_all

Ultimately, I recommend choosing the option "Export all normalized intensities" and looking at the expression values for a few of those probesets in the files:
1) ExpressionInput/exp.YourDataset.txt
2) ExpressionInput/stats.YourDataset.txt
3) AltExpression/ExonArray/Hs/YourComparison
4) AlternativeOutput/AltResults/RawSpliceData/Hs/splicing-index/YourComparison
5) AlternativeOutput/DomainGraph/YourComparison

The first will show the probeset RMA expression values, the second the DABG p-value, third the expression values after filtering (for those that align to Ensembl genes and that meet your expression filtering options), fourth the normalized intensities for each probeset and five the splicing-index or FIRMA scores along with associated p-values for all probesets. This will help you track down what is going on.

Best,
Nathan

--
You received this message because you are subscribed to the Google Groups "Alternative Splicing and Functional Prediction" group.
To post to this group, send email to alt_pre...@googlegroups.com.
To unsubscribe from this group, send email to alt_predictio...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/alt_predictions?hl=en.