[BioC] edgeR with two different groups (unpaired) at two time points (paired): Need some input

Sindre sindre.lee at studmed.uio.no
Thu Oct 24 10:34:35 CEST 2013


Thank you so much.

If genes whose response differs in diabetic patients vs healthy = 0.

Would it make any sense to "merge" the groups, thus only diabetics vs healthy 
or time point 1 vs time 
point 2, to find differences there?

And also, just to be sure, is this correct?

d <- (calcNormFactors(d,method="RLE"));
d <- estimateGLMTrendedDisp(d, design, verbose=TRUE)
d <- estimateGLMTagwiseDisp(d, design)
fit <- glmFit(d,design)

Should I use other than default degrees of freedom (I have 48 samples)?



More information about the Bioconductor mailing list