Comparing Results from featureCounts and htseq_counts

Question

[oma219]

Hello,

I was wondering if there any typical way people go about comparing the results they get from feature counts and htseq_counts? I wanted to make sure that they produced similar results but when I looked at some individual genes they didn't match that closely. Thanks

Answer 1

[Devon Ryan]

If you use the same settings with both, then their concordance is somewhere around 99% (the difference used to be due to an off-by-one error in htseq-count, no clue if that's still an issue). What you're likely seeing are the results of different settings (whether default or otherwise).

[oma219]

Oh okay, do these two commands have the same settings for the programs: featureCounts: featureCounts -g Name -a /home/ubuntu/data/rnaseq/nematostella/venus/genome/Nemostella_New_Annotation.gff3 -o Venus4_counts_htseq.txt /home/ubuntu/data/rnaseq/nematostella/venus/alignment/Venus4_STARAligned.out.sam

Htseq-Count: htseq-count -i Name -s no /home/ubuntu/data/rnaseq/nematostella/venus/alignment/Venus4_STARAligned.out.sam /home/ubuntu/data/rnaseq/nematostella/venus/genome/Nemostella_New_Annotation.gff3 > Venus4_counts_htseq.txt

Also, when you look at the featureCounts output, the counts of each gene is the last column correct?

[Devon Ryan]

I'm not going to bother going through all of the defaults for each program and comparing them, I'll let you do that.

Yes, I recall that the last column is the one you want (well, last columns, since you can have featureCounts process multiple samples at once).