Plotting Coverage - Strand specification

Question

[Constantine]

Hello

I'm trying to create a coverage plot from a bam file (H3K4me3) via NGSplot and/or Deeptools. This is the output from both softwares

I understand the peaks I'm getting at the TES are because I'm not specifying strand information.

I looked at the manuals for both NGSplot and Deeptools and didn't find anything about strand specification

Any idea how to circumvent the problem?

Thanks

[Chirag Nepal]

It may be due to the strand of the genes you are plotting.

To check, divide your genes into + strand and - strand, plot as before. You should see peak on 5' end only on genes in + strand, and peak on 3' end only on genes in -ve strand.

Answer 1

[Brian Bushnell]

You can generate strand-specific coverage from bam files with the BBMap package like this:

pileup.sh in=mapped.bam basecov=cov_#.txt strandedcov=t

You can then plot it with R or something.

Oh... also there's a script called splitsam.sh that will split a sam file into 3 sam files, one with plus-mapped reads, one with minus-mapped, and one with unmapped. That could be useful.