How to filter rRNA reads in SAM file.

[zee]

I would stick with all the reads mapped to the genome and just filter out the regions you dont want. You could accomplish this task without writing a single script using the BEDtools package.

BEDtools accepts BAM-formatted files. You would need to place all your rRNA regions in BED format e.g. a simple illustration of 2 regions in a file called regions.bed

chr1 1 10 rRNA1
chr2 2 2333 rRNA2

So if your reads are in a file called alignments.bam Then do

intersectBed -abam alignments.bam -b regions.bed -v > filtered.bam

Note I use the "-v" option to find regions in file a that DOES NOT intersect with b.

If you're not comfortable doing this on the command line I suppose you could upload your BAM file to Galaxy and do the same operation with intersectBed.
Hope this helps.

[townway]

I am not a good script writer, so is there any tools available to deleted these reads?

Thank you

[epigen]

To my knowlegde, unfortunately no. You could try mapping your reads to rRNA sequences and exclude those that align or, if you already know the genomic region (as it seems), just ignore reads at these coordinates, e.g. by writing a script to kick them out of your SAM file.

[townway]

a lot of reads in my data were mapped to rRNA region, how can I filter them in SAM format. is there any command in samtools or other package?
Thanks.