Hi, 

csaw allows one to remove PCR duplicates when doing read counting across windows. My question is how is this functionality actually executed.

I generally mark duplicates using Picard Tools or Samblaster and I was of the view that csaw would remove these reads when doing read counting. However, I was wondering if csaw calls some other program (or its own method) internally for marking and removing duplicates. 

It would be great to get some insight into this. 

Thank you, 
Best, 

Asif 

I think section 2.2.2 of the csaw User's guide covers this pretty clearly. In addition, ?readParam has this as the first (!) paragraph under 'Removing low-quality or irrelevant reads':

     Marked duplicate reads will be removed with  dedup=TRUE . This may
     be necessary when many rounds of PCR have been performed during
     library preparation. However, it is not recommended for routine
     counting as it will interfere with the downstream statistical
     methods. Note that the duplicate field must be set beforehand in
     the BAM file for this argument to have any effect.