USeq - FilterDuplicateAlignments - SAM

[alperyilmaz]

I am trying to use USeq for ChIP-Seq analysis. I wanted to run FilterDuplicateAlignments just in case there is potential amplification bias in the sample. Guess what, formatting(1) problem!!

I used Bowtie to align and output is in SAM format. USeq accepts SAM, there's no problem there. The problem is FilterDuplicateAlignments accepts a tabular output and filters the alignments those with the same chromosome, position, and strand. You tell which columns contain chromosome, position and strand and the application takes care of the rest. However, SAM does not keep strand information in a human readable info.

I was wondering if it's okay to add an additional column in SAM output which contains "+" or "-" so that FilterDuplicateAlignments can parse it? If I temper with SAM output will that effect downstream applications?

As some of us know, tempering with SAM output is also needed if TopHat and Cufflinks are used together (1,2)

thanks..


(1) I wonder what percent of the encountered problems or questions being asked in this forum are solely related to formatting issues..
(2) Cufflinks need strand information in an custom column in SAM output. But TopHat output does not contain that custom column. You need to manually temper with TopHat's output

[bioinfosm]

I am curious about what you discovered!

I am trying to use USeq for ChIP-Seq analysis. I wanted to run FilterDuplicateAlignments just in case there is potential amplification bias in the sample. Guess what, formatting(1) problem!!

I used Bowtie to align and output is in SAM format. USeq accepts SAM, there's no problem there. The problem is FilterDuplicateAlignments accepts a tabular output and filters the alignments those with the same chromosome, position, and strand. You tell which columns contain chromosome, position and strand and the application takes care of the rest. However, SAM does not keep strand information in a human readable info.

I was wondering if it's okay to add an additional column in SAM output which contains "+" or "-" so that FilterDuplicateAlignments can parse it? If I temper with SAM output will that effect downstream applications?

As some of us know, tempering with SAM output is also needed if TopHat and Cufflinks are used together (1,2)

thanks..


(1) I wonder what percent of the encountered problems or questions being asked in this forum are solely related to formatting issues..
(2) Cufflinks need strand information in an custom column in SAM output. But TopHat output does not contain that custom column. You need to manually temper with TopHat's output