Hi all. I am using BWA mem to map reads from a ChiP experiment which comes from an F1. To avoid mapping bias we are mapping them to a diploid genome consisting of both parent genomes. To get unique reads we filter them by q20 but we also want to recover all the multi-map reads which map twice, once to the father and once to the mother genome. We do not want to lose all those positions which are TF binding site, there is just no variation at the site. Since we are doing this for many F1s losing all those non-variant sites would give us very biased results.
I know MAPQ=0 is ambiguous reads but is there a way to distinguish reads that map twice and reads that map 3 or more times with equal score?
Your help would be appreciated.