How does cBioportal calculate p-values in the 'co-expression' section?

[wcharle...@gmail.com]

Hello,

After doing a few rounds of testing on my own to see if I could recapitulate regression coefficients that cBioportal produces, I found that the tool uses Spearman correlation and used the 'median expression' data from TCGA (not normalized z-score). Next I want to get the same p-value. 

Which method is used to obtain a p-value, and is some correction (fdr, Bonferonni, etc) applied ? 

Thanks!

Charlie

[adamab...@gmail.com]

Hi Charlie,

Thanks for your question. The code which calculates the p-value based on the correlation can be found here, in the function `computeCorrelationPValue` 

https://github.com/cBioPortal/cbioportal-frontend/blob/master/src/shared/components/plots/PlotUtils.ts#L233

Basically it's a 2-sided t-test, with a t-statistic derived from the correlation.

Thanks,

Adam

[JJ Gao]

Thanks, Adam.

Adding Charlie to the email and add that the q-values are adjusted from p-values with Benjamini-Hochberg FDR correction procedure.

-JJ

--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
To post to this group, send email to cbiop...@googlegroups.com.
Visit this group at https://groups.google.com/group/cbioportal.
To view this discussion on the web visit https://groups.google.com/d/msgid/cbioportal/c60ca62f-55ce-4743-995d-50ddbd915348%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.