Hello there, 

I am a little confused by the toptable output (well really the FDR) and thought I'd quickly check my interpretation. 

For the following code:

topall <- toptable(fit, coef="CaseAD", genelist=fit$genes, number=Inf, adjust="fdr", p.value=0.01, lfc=log2(1))

Would the interpretation be:

Using a threshold of 0.05 for FDR adj. p.values and a minimal fold change of 2, this analysis produces n number of genes?

And does the above mean that I would look at the adjusted p.values? 

I am interested in following MAQC guidelines for this analysis, so would rather restrict my results to a FDR = 0.05 and p.values < 0.01. However, I am not entirely sure, how to get that, since I am not sure what the "adjust=fdr" relates to here? 

Any advice?

Many thanks!!!

Your topTable call is asking for all genes that have FDR below 0.01 and log-fold changes greater than log2(1), i.e., zero. This effectively means that you're only selecting genes based on the FDR. Which is totally fine, because you should always be using adjusted p-values for genomic analyses.

I don't know what the MAQC guidelines say, but setting thresholds on both the FDR and p-values seems pointless to me; the former is calculated from the latter, so they'll be redundant. I would recommend just setting a FDR threshold (usually 1 or 5%) and defining all DE genes as those with FDR values below that threshold.

If you want to include a log-fold change threshold, use treat rather than setting it in topTable; see the warnings at the bottom of the documentation for ?topTable.