When is HTSeq Count Script used?

Question

[oma219]

Hello,

I was just wondering when in the pipeline is the HTSeq count function used? Is it only used when DESeq is being used for further analysis? Thanks.

Omar

Answer 1

[Benn]

HTSeq can be used for any DE tool that uses raw read counts (hence limma trend, edgeR, DESeq2). You can also try featureCounts instead of HTSeq, which I personally prefer.

[oma219]

Does feature counts have any additional features over HTSeq? Thanks!

[gabriel.rosser]

1) HTSeq seems to have a bug that causes it to fail with certain libraries: this thread describes a bug that I personally have run into. Note that the thread ran for almost 2 years with no robust fix.

2) featureCounts puts all the results into a single output file, making it slightly easier to load (I don't know if htseq-count supports that, it wasn't obvious to me anyway)

3) featureCounts is faster, according to its authors. See Table 1 in their open access paper. I have also found this, but haven't timed it properly.

[GenoMax]

featureCounts also supports SAM v.1.4 format tags (not to be confused with version of samtools program) and will automatically sort your BAM files as needed. Last time I had looked htseq-count did not do either.

featureCounts is significantly faster than htseq-count as has already been mentioned.

[Benn]

I am not sure, I found it easier in usage...

[WouterDeCoster]

Comment moved to an answer ;-)