Hi all, I was wondering how the p-value correction using fdr works. The following Limma package’s command “ topTable(fit,number=100, adjust="fdr")” corrects the p-values but I don’t know which is the percent of false positive we can expect using this correction. Is the FDR fixing at one pre-determined value, e.g. 5 %? Thank you in advance Best, Manuel --------------------------------- [[alternative HTML version deleted]]

See ?p.adjust for specifics. However, in a particular case, if one has 100 genes with fdr less than 0.05, then 5 out of that 100 are expected to be false positive. In other words, for any given fdr cutoff x, if there are N genes with fdr<x, then="" x*n="" genes="" are="" expected="" to="" be="" false="" discoveries.="" sean="" on="" dec="" 10,="" 2004,="" at="" 7:21="" am,="" m="" perez="" wrote:=""> > Hi all, > > I was wondering how the p-value correction using fdr works. > > The following Limma package's command " topTable(fit,number=100, > adjust="fdr")" corrects the p-values but I don't know which is the > percent of false positive we can expect using this correction. Is the > FDR fixing at one pre-determined value, e.g. 5 %? > > Thank you in advance > > Best, > > Manuel > > > > --------------------------------- > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor

Dear Manuel, You are selecting the top-ranked number genes, ordered using the "sort.by" option; the p-value you are seeing is the fdr-adjusted p-value, so no, no pre-determined cut-off has been selected. Some explanation is provided in the help for p.adjust; here you also have an extensive list of references. You might want to add to that the documentation of the multtest package which provides a quick overview of multiple testing issues, including the definition of adjusted p-value according to FDR. Best, R. On Friday 10 December 2004 13:21, M PEREZ wrote: > Hi all, > > I was wondering how the p-value correction using fdr works. > > The following Limma package?s command ? topTable(fit,number=100, > adjust="fdr")? corrects the p-values but I don?t know which is the percent > of false positive we can expect using this correction. Is the FDR fixing > at one pre-determined value, e.g. 5 %? > > Thank you in advance > > Best, > > Manuel > > > > --------------------------------- > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor -- Ram?n D?az-Uriarte Bioinformatics Unit Centro Nacional de Investigaciones Oncol?gicas (CNIO) (Spanish National Cancer Center) Melchor Fern?ndez Almagro, 3 28029 Madrid (Spain) Fax: +-34-91-224-6972 Phone: +-34-91-224-6900 http://ligarto.org/rdiaz PGP KeyID: 0xE89B3462 (http://ligarto.org/rdiaz/0xE89B3462.asc)