Samtools View Of A Sam File Entire Chromosome

Question

[dfernan]

Hi,

I have (what I think) is a valid sam file.

When I do:

samtools view -S F1i1F.C57.2.Aligned.out.sam chr19_C57

I get the follong message:

[samopen] SAM header is present: 66 sequences.

[main_samview] random alignment retrieval only works for indexed BAM files.

Note: my chromosomes are chr1_C57, chr2_C57, etc.

As far as I remember I was able to see sam files by chromosomes when using an unsorted sam file but I may be wrong. I can transform it to bam, sort it and index it, but is it necessary?

Answer 1

[Zev.Kronenberg]

You must index your bam file. Querying works by looking up where the seqid is in the index. You must sort and index.

Answer 2

[Devon Ryan]

As far as I remember I was able to see sam files by chromosomes when using an unsorted sam file but I may be wrong. I can transform it to bam, sort it and index it, but is it necessary?

Yes, using only samtools view, you'll need a sorted and indexed BAM file. However, you can also just use grep:

grep -w chr1_C57 F1i1F.C57.2.Aligned.out.sam

To get the reads on that chromosome. That will miss the header, of course. I can conceive of a few edge cases where this wouldn't work (namely, putting the chromosome in an auxiliary tag that also contains spaces), but you could just use a simple awk command if that's the case.

[Ashutosh Pandey]

grep -P "\tchr1_C57\t" F1i1F.C57.2.Aligned.out.sam should work.

egrep '^@|chr1_C57' F1i1F.C57.2.Aligned.out.sam should get you the header too. This will only give meaningful results if the readname doesnt start with "@"

[Devon Ryan]

Good call, the regex is definitely a better method than using word match.