Obtaining strand calls from BAM/SAM

[Protaeus]

I am interested in extracting strand info at any or all given positions from an alignment. The purpose is to get info about strand bias in a haploid organism. I know GATK provides strand bias in their genotyper output, but I want the actual counts...for example, at position 500 in my reference, 250 are derived from one strand and 250 are from the other. I'm not concerned about the call, per se, rather the value associated with the calls at that position.

If this makes sense to anyone and you have some ideas or a tool that already can accomplish this task, let me know.

Thanks...

[swbarnes2]

An all-points vcf file from samtools will output this. The first two numbers of the DP4 will tell you how many reads fell in the forward direction, and the reverse direction, and matched the reference. The third and fourth numbers tell you how many fell in the forward and reverse directions that didn't match reference, but you might not care about those.

I'd make sure that the people prepping your sample actually did a prep that will preserve strand information, as standard preps for most applications won't.