RNA-Seq: GFOLD counts is different from HTSeq-count

[super0925]

Does anybody use GFOLD before?

http://compbio.tongji.edu.cn/~fengjx/GFOLD/gfold.html

Why do I use these two tools count reads but give different results?

In GFOLD manual, “Because of possible overlapping of multiple genes, a read could be mapped to the overlaped region of multiple genes. In this case, a read is counted multiple times with each time for each gene. Furthermore, if a gene is on multiple chromosomes or different strands of the same chromosome, only exons on one strand of one chromosome (the one appear first in the annotation file) will be assigned to this gene. Exons not on this strand of the chromosome will be discarded.”

gfold count -ann genes.gtf -tag sample1.sam -o sample1.read_cnt

htseq-count --minaqual=10 --mode union sample1.sam genes.gtf>htseq.counts

Is that because HTSeq skip the alignment quality lower reads?

[dpryan]

The part you showed from the GFOLD manual is completely different from what htseq-count and featureCounts do, so it's expected that you'd get different results.