This is a test version of Biostars. For the public version, visit https://www.biostars.org.
How to split SAM file to different chromosomes

I have a SAM file which I now want to split into different chromosome from chromosome chr1..chr20 as follows.

/home/SpliceGraph/Human/chr1.sam
/home/SpliceGraph/Human/chr2.sam/
/home/SpliceGraph/Human/chr3.sam/
/home/SpliceGraph/Human/chr4.sam/
/home/SpliceGraph/Human/chr5.sam

Which command can I use to do this?

assembly rna-seq
• 2,207 views • link
updated by Biostar Biostar • written by javanokendo • javanokendo

The following code: samtools idxstats out.bam | cut -f1 | grep -v '*' > chr.names is not giving the list of chromosomes. It give something like this:

NC_000001.11
NT_187361.1
NT_187362.1
NT_187363.1
NT_187364.1
NT_187365.1
NT_187366.1
NT_187367.1
NT_187368.1
NT_187369.1
NC_000002.12
NT_187370.1
NT_187371.1
NC_000003.12
NT_167215.1
NC_000004.12
NT_113793.3
NC_000005.10
NT_113948.1
NC_000006.12
NC_000007.14
NC_000008.11
NC_000009.12
NT_187372.1
NT_187373.1
NT_187374.1
NT_187375.1
NC_000010.11
NC_000011.10
NT_187376.1
NC_000012.
• 0 views • link
updated by GenoMax GenoMax • written by javanokendo • javanokendo

Those are the chromosome (reference) names that are present in your BAM file.

Please use ADD COMMENT/ADD REPLY when responding to existing posts to keep threads logically organized. SUBMIT ANSWER is for new answers to original question.

• 0 views • link
written by GenoMax GenoMax

Please use ADD COMMENT to reply to answers. It gives you the names that were in the fasta file that you mapped against. The command itself is correct. If you want different names then align against a fasta file that contains the names you would like.

• 0 views • link
written by ATpoint ATpoint

1 answer 

## Convert SAM to BAM and index it:
samtools view -o out.bam in.sam
samtools index out.bam

## Extract chromsosome names:
samtools idxstats out.bam | cut -f1 | grep -v '*' > chr.names

## Split bam file with w while loop
while read p
  do
  samtools view -o out_${p}.bam out.bam ${p}
  done < chr.names

If you really want SAM instead of BAM files then use samtools view -ho out_${p}.sam out.bam ${p}. Given you have the resources you can of course use something like GNU parallel instead of a loop to make it more efficient.

• 0 views • link
written by ATpoint ATpoint

Log in to answer this question.

More posts like this