Hi Sabrina.
This is really a limma question, not an aroma.affymetrix one, but I'll
take a quick stab. Basically, you'll notice that regardless of the
value you set for 'p', you'll get the same t-statistics and p-values.
It only changes the B values (posterior log odds).
Cheers,
Mark
On
Hi, all:
A quick question. I know after we extract frima score, I should use
eBayes to do the following:
assume we have design matrix and contrast matrix,
fit<-lmFit(exFirma,design)
fit2<-contrasts.fit(fit,cont.matrix);
fit2<-eBayes(fit2,proportion=p);
the default setting for p is 0.01 that is fo
