How to remove reads associated with a specific region from bam file?

Question

[noushin.farnoud]

Dear Members,

Is there a way I can removes reads associated with a region (chr, start, end) from a .bam file (RNASeq data) prior to the application of HTSeq?

I will greatly appreciate your feedback

Noushin

Answer 1

[Sergey Naumenko]

bedtools intersect -abam file.bam -b filter.bed -v > filtered.bam

filter.bed should contain

chr    start     end

[abaluapuri]

Thats perfect solutions !! Super cool ! thanks !!!

[LDT]

Do you need to specify the positive and negative strand as well?

Answer 2

[sunhanice]

Just found, there is an option -U in samtools view. Use it like this:

samtools view input.bam -b -h -o output_inRegions.bam -U output_outRegions.bam -L Regions.bed

[varsha619]

Just to clarify, I used

samtools view in.sorted.bam -b -h -o inRegions.bam -U outRegions.bam "chr:start-stop"...

So here the -o file excludes the regions chr:start-stop and has the rest, but the -U file only retains the chr:start-stop? Thank you for your help!

Answer 3

[John]

You'll want to use NGSUtil's bamutils tool, specifically with -excludebed.

But, I'd recommend you don't :P

The BAM format is to store highly compressed alignment data. You should treat them like raw, virgin data, without normalization/filtering tweaks here and there to get it into shape.

All that kind of intersection stuff should be done on processed signal data - wigs and bedgraphs, etc - where its much easier to have multiple versions of things and to just dump it all and start afresh from the .bam if you have to.

Having said that, its your data, do what you like with it :)

Answer 4

[igor]

If you are filtering your BAM for HTSeq, then you are doing extra work. You should just modify the GTF file that you are giving to HTSeq to exclude regions you do not want.

Answer 5

[Ron]

Using this QC package for RNAseq: http://rseqc.sourceforge.net

Split_bam.py would do the splitting of bam files.