Dear List, I have a question regarding p-values, d-values and q-values returned by Bioconductor's SAM method (siggenes package). - My understanding is that the p-values returned are unadjusted. Does it make sense to apply standard FDR procedres (e.g. Benjamini- Hochberg) to them? - What is the definition of d-values and q-values? Best, Michal Blazejczyk

Hi Michal, the d-values are the values of the moderated t-statistic (called d-statistic in the Tusher paper about SAM). The q-values are the FDR adjusted p-values as proposed by John Storey. And so yes, you can (and actually should if you wanna use them) adjust the raw p-values with any multiple correction procedure. The only difference of these p-values (when computed permutation-based) to other permutation-based p-values is that the p-values in SAM are computed based on all mB permuted values for the m genes and B permutations, while the usually used p-values are determined on the B gene-specific permutations. Best, Holger -------- Original-Nachricht -------- > Datum: Wed, 10 Nov 2010 13:25:27 -0500 > Von: Michal Blazejczyk <michal.blazejczyk at="" mail.mcgill.ca=""> > An: bioconductor at stat.math.ethz.ch > Betreff: [BioC] SAM results > Dear List, > > I have a question regarding p-values, d-values and q-values returned > by Bioconductor's SAM method (siggenes package). > > - My understanding is that the p-values returned are unadjusted. > Does it make sense to apply standard FDR procedres (e.g. Benjamini- > Hochberg) to them? > > - What is the definition of d-values and q-values? > > Best, > Michal Blazejczyk > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor --