Confusion in HTSeq paired-end output count

Question

[Whoknows]

Hi friends

I used HTSeq for creating count table (outputs) in paired-end file, actually I generated RPKM values from these count tables with below formula :

• C = Number of reads mapped to a gene
• N = Total mapped reads in the experiment
• L = exon length in base-pairs for a gene
• Equation = RPKM = (10^9 C)/(N L)

Is it right to use exact read count numbers of the count tables for RPKM ??

I mean, because they are paired-end files, this equation don't need to be correct by dividing values (C and N) on 2? OR not? OR HTSeq consider this issues too?

Thanks

Answer 1

[Devon Ryan]

Technically you're getting FPKM rather than RPKM, but since htseq-count isn't going to count reads whose mate is aligning to a different feature anyway they should be the same (i.e, the counts would be doubled but so would the total number of mapped reads). To allay what's likely your biggest fear, the numbers produced by htseq-count when paired-end reads are used relates to the number of pairs or singletons rather then to individual reads, so there's no division by 2 needed