Dear Michael,

I noticed a very small type in the Time Course section of the RNA-Workflow (http://bioconductor.org/packages/release/workflows/vignettes/rnaseqGene/inst/doc/rnaseqGene.html#time-course-experiments). The 7th code part shows

res30 <- results(ddsTC, name="strainmut.minute30", test="Wald")
res30[which.min(resTC$padj),]

I think it should be which.min(res30$padj) and not resTC

Now to my question. The vignette suggest to use alpha and filterFun=ihw for the results() function (e.g. when I do a comparison A vs B). Is this recommend for the time course as well? because these options are not mentioned in time course part of the vignette.

For the next I use the chunks and parts of the fission data set because my design is identical. I know that

res30_w <- results(ddsTC, name = "strainmut.minute30", alpha = 0.1, test = 'Wald')

will provide me with Mut vs WT at a time point 15, controlling for any difference that existed for that Mut compared to WT at time 0h. And thats what we are interested in. But we are also interested in genes that change up to a certain timepoint. But

ddsTC <- DESeqDataSet(fission, ~ strain + minute + strain:minute)
ddsTC <- DESeq(ddsTC, test="LRT", reduced = ~ strain + minute)
resTC <- results(ddsTC)

will only provide me with a list of DEGs over all timepoints. Is it possible to get DEGs up to e.g. minute 30 or minute 60?

If I run

res30 <- results(ddsTC, name = "strainmut.minute30", alpha = 0.1)
res60 <- results(ddsTC, name = "strainmut.minute60", alpha = 0.1)

then the results tables are identical only the log2FC is different. What is the log2FC representing here?

Thanks for your Time!

Thanks for pointing out the typo! I've pushed a fix to devel.

IHW can be beneficial in some cases over the standard filtering. I think if you are interested, and it's easy for you to add to your pipeline, it's worth it.

Is it possible to get DEGs up to e.g. minute 30 or minute 60?

You can get this by taking the union of individual sets that are FDR significant for individual tests.

the results tables are identical only the log2FC is different

You can generate Wald test p-values with test="Wald", otherwise it is giving you the LRT p-value. See description of test in ?results.