Stranded BedGraph/BigWig for paired-end reads

[id0]

For stranded data, I use "bedtools genomecov" with -strand to generate coverage plots. For single reads, this works fine, but for paired-end reads, it doesn't actually take mate pairs into account. Thus, half the data (reverse reads) ends up on the "wrong" strand. Is there a way to generate coverage plots for paired-end stranded data?

Is the only solution to split the BAMs, process them separately, and then merge? That seems like a lot of work. Also, very heavy on the I/O.

[biocomputer]

I am about to start RNA-seq and I'm wondering the same thing. When it comes time to generate coverage files from the aligned sam/bam files, will bedtools genomecov (using the -split option) correctly process paired-end sequencing? If not is there another method to generate these coverage files?

[volks]

you can use igvtools count --strands first for creating a wig file that contains columns for both strands. subsequently use cut/awk and wigToBigWig.

[yoyoq]

since this is one of the top google hits for this problem.
let me plug

deeptools (the package)
and
bamCoverage (the command)

it seems to do it without the gd flags and without all the awk