Split bam by multiple chromosome to a single bam file

Question

[l0o0]

Hi, I am using bowtie + samtools pipeline to call snp. Split bam file and call snp by chromosome will save a lot of time.

But the reference genome have many scaffold, split bam by chromosome will produce a lot of scaffold_bam file. Now I want to split bam by scaffold, so the scaffold will split into one file.

Is there any way to do that?

I tried to use command:

samtools view in.bam scaffold1 scaffold2 -b > scaffold1_2.bam

but I don't know how to check if scaffold1_2.bam contains scaffold1 and scaffold2.

Thanks

[l0o0]

I've tried to do the command above, the output contains 2 scaffolds!

samtools view in.bam scaffold1 scaffold2 -b > scaffold1_2.bam

it works

Answer 1

[dariober]

Maybe something on these lines?

First prepare a string of required scaffold names. You can extract all the scaffold names for the bam header and use grep to get only those matching a certain patter. In this example get only scaffolds starting with "chr1":

chroms=`samtools view -H in.bam \
| awk '$1 == "@SQ" {sub("SN:", "", $2); print $2}' \
| grep -P '^chr1.*'`

Then pass this string to samtools. If the string is really long, you might need xargs to split it otherwise you exceed the maximum length of a single command (assuming you are on *nixsystem):

echo $chroms | xargs samtools view -b in.bam > scaffold1_2.bam

(Not fully tested...)

[l0o0]

Hi dariober. I've tested your commands, and it works. Thanks for your replay.