Counting Reads in SAM Format

[foolishbrat]

Typically mapping output will show many reads that are very close to each other. Meaning the difference of mapping location differ within 10 positions.

Is there any existing tool to cluster such reads from SAM output and compute the frequency of reads mapped in the cluster?

Surely we can write an all-against-all read comparison, but it takes very longtime to run.
Perhaps there are faster tool around to do this task?

[francois.sabot]

Hi,
you can probably sort them by coordinatées using the Picardtools Samtools, and then using home-made scripts or an alternate use of Depth-of-Coverage to determinate the positions with such high level clusters (high density).

[berath]

HTseq-count might answer your needs:

http://www-huber.embl.de/users/anders/HTSeq/doc/count.html