Hi,

I am trying to find the DE genes between various contrasts using limma. My experimental design include 10 samples before immunosuppressive therapy and same 10 samples after 12 weeks of therapy and 24 weeks of therapy. I am trying to find the DE genes between various conditions like T12-T0, T24-T0.

I took support of limma userguide and I designed and created my contrast matrices. My question is about topTableF. I understood that topTableF is F statistic. Does it give the lfc values. I am unable to interpret my result correctly. I would like to see how my DE genes are varying in each condition. I am using the following code.

library(limma)
# limma
design <- model.matrix(~0 + eset_filtered$cohort)
colnames(design) <- levels(eset_filtered$cohort)
fit <- lmFit(eset_filtered, design)
contrast.matrix <- makeContrasts("T12-T0","T24-T0",levels = design) # make sure your column name is matched to cohort name
fit2 <- contrasts.fit(fit, contrast.matrix)
fit2 <- eBayes(fit2)
top.active.1 <- topTable(fit2, coef=1, adjust="BH", n=nrow(eset_filtered)) # 23,907
top.active <- topTableF(fit2, adjust="BH",n=nrow(eset_filtered))
dev.off()
plot(density(top.active$adj.P.Val), main="Probe density of LIMMA BH corrected p-vals")
index <- which(top.active$adj.P.Val < 0.2)
top.active.sig.one <- top.active[index,]

My output looks like this. CAn I consider the T12.T0 and T24.T0 as logFC values. And also can I get the differentially expressed genes based on F statistic. Could anyone please help how to interpret this result. The main aim is I need to find DE genes changing over time before and after treatment. Any helpis much appreciated.

SYMBOL T12.T0 T24.T0 AveExpr F P.Value adj.P.val

C5orf63    -0.906  -0.9213  3.70     16.7           0.0847

Thanks

CAn I consider the T12.T0 and T24.T0 as logFC values.

Yes

And also can I get the differentially expressed genes based on F statistic.

You already have - that's what topTableF does.

Could anyone please help how to interpret this result.

An F-statistic tests the hypothesis that one or more of the contrasts is different from zero (or more correctly, the result of computing one or more of the contrasts...). So for any of the genes in your topTableF result, you have evidence that the gene is differentially expressed in one or more of the comparisons.