Subset Bam File According To List Of Contigs

Question

[patr.grossmann]

Dear Biostar community,

I have a bwa mapping of metagenomic sequencing reads on assembled contigs. I also have a list of IDs of "interesting" contigs. Now I would like to subset the mapping (bam/sam file) to obtain a mapping (bam/sam file) that only contains my desired contigs. Do you know of any simpel way (samtools, picard tools, R, Scala/Java) to do so? Note please that creating a mapping with a reduced contig.fasta file is not an option for me.

Thank you in advance!

Answer 1

[Ashutosh Pandey]

samtools view -bh input.bam (should be sorted) contig_of_interest > output.bam

For example: samtools view -bh input.sorted.bam chr1 > output.bam . In case your contig of interest is named as chr1.

if you have more than one contig of interest then go for:

samtools view -h input.sorted.bam | awk '{if($3 == "chr1" || $3 == "chr2"){print $0}}' | samtools view -Sb - > output.bam. In case your contigs of interest are chr1 and chr2.

[patr.grossmann]

Thank you both for your answers. :) May I add that you can just append the regions: So one could do samtools view -bh input.sorted.bam chr1 chr2 > output.bam or samtools view -bh input.sorted.bam cat list_of_contigs.txt | tr "\n" " " > output.bam

[Ashutosh Pandey]

Thanks for the additional information. I didn't know that.

[lagartija]

I know it's old but for new arrivers this is what worked for me : cat chromosome.list | tr "\n" " " | xargs samtools view -bh input.bam > output.bam

[Ashutosh Pandey]

If you havent converted your sam files to bam then use :

egrep "^@HD|^@RG|^@PG|chr1|chr2" input.sam > output.sam (filtered)