Hello, I am trying to work out the meaning of logFC in the toptable from limma package. Some sources say that it is the log2 of FC and others say that it is not. Is this value the actual difference in averages of cases and controls (and as such can it be used to determine effect size)? Regards, Jack -- output of sessionInfo(): R version 2.15.2 Patched (2012-11-28 r61176) -- Sent via the guest posting facility at bioconductor.org.

Hi Jack, On 8/9/2013 5:02 AM, Jack [guest] wrote: > Hello, > I am trying to work out the meaning of logFC in the toptable from limma package. Some sources say that it is the log2 of FC and others say that it is not. The logFC column is a coefficient from your model (or the numerator from a contrast). If you have set things up correctly, it will be the log2 of FC between two or more sample types. In other words, if you are doing a simple case vs control, and you specify the model correctly and select the proper coefficient for your topTable(), then the logFC will be the log fold change between the two groups. And note that log(A/B) = log(A) - log(B), and what you are doing with limma is usually taking logged data and computing the difference in means. So you are computing the right hand side of that equation, but it is identical to the left hand side. Best, Jim > > Is this value the actual difference in averages of cases and controls (and as such can it be used to determine effect size)? > > Regards, > Jack > > -- output of sessionInfo(): > > R version 2.15.2 Patched (2012-11-28 r61176) > > -- > Sent via the guest posting facility at bioconductor.org. > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor -- James W. MacDonald, M.S. Biostatistician University of Washington Environmental and Occupational Health Sciences 4225 Roosevelt Way NE, # 100 Seattle WA 98105-6099