how to filter a bam file

[paolo.kunder]

Dear All,
I am facing a problem with huge rRNA and globin RNA contamination in my SOLiD data (colour space format csfasta)
I would like now to get rid of this reads in my csfasta files or in my bam file ( I used tophat to align).
I s there a way to delete those reads from either the files?
Any suggestions will be appreciated,
Paolo

[arvid]

I like to do it with BEDTools; I use "intersectBed -v" to remove the reads mapping to rRNA regions (given in a GFF file).

First get the GFF with rRNA regions:

grep -P "\trRNA\t" annotation.gff > rRNAs.gff

You might need to adapt the matching pattern here, works fine with e.g. Arabidopsis TAIR10 GFF3.

Then get all reads not intersecting with these regions:

intersectBed -v -abam tophat_aligned_reads.bam -b rRNAs.gff > tophat_aligned_reads_no_rRNAs.bam

[paolo.kunder]

thanks for your help I get rid of my junk sequences!