I have exome-sequencing from mouse tumors as well as constitutional DNA from the tails (Illumina, 50bp paired end). Tumor sequencing data looks good, but the data from the tails does not. One thing in particular is the percentage of reads with a mate pair that maps to a different chromosome. In tumors this percentage is around 0.2%, while in the tail samples it is around 15%--that is, 15% of the reads that have a paired read that mapped to a different chromosome. I am going to investigate this further, but I wanted to know if anyone has anyone else encounter a similar problem? I' am thinking to just simply remove those 15% of the reads from the tail samples.