Bam File: Extract Chromosome Number And Start Position Reads

Question

[Lisanne]

Dear All,

I am working with BAM files for chip-seq analysis. For each read in the BAM file, I want to extractt the chromosome number, the start position and the stop position and replace that in a new file. Is that possible?

Many thanks!

Answer 1

[Farhat]

samtools view bamfile.bam|awk '{print $3 "\t" $4 "\t" $4+length($10)-1}' > newfile.tab

will do the job. The stop position here is the last matching position.

[Aaronquinlan]

bedtools' bamToBed inspects the CIGAR string when computing the end coordinate, so deletions are properly handled. the example here assumes that only substitutions can occur,

[Wen.Huang]

this approach has a minor issue that the length of the sequence does not necessarily agree with the span of the alignment, e.g. indels

[Lisanne]

Thanks to Farhat, that works fine for me! :)

[Farhat]

True. bamToBed would be the proper tool to handle something like this.

Answer 2

[Michael]

Yes it is pssible, using e.g. SAMtools samtools view or Rsamtools.

The reference sequence name is in column 3 of a SAM file, the (leftmost) start in column 4 and the end position needs to be calculated using the CIGAR string (e.g. start + alignment length).

[mjg]

In this case how is strand information treated? Is it always start + alignment length or does it depends on the strand where the read mapped?

Answer 3

[Swbarnes2]

You could also try BEDTools, which can convert your .bam to a .bed, and bed pretty much is a name, a chromosome, a start and a stop position.