I a m running DiagnoseTargets on a list of intervals corresponding to targeted exons. In the result file, intervals are filtered by i.e. PASS, LOW_COVERAGE, COVERAGE_GAPS or NO_READS for each sample as well as for the sample set as a whole. In the individual-sample FORMAT fields, information on TF (filter) and IDP (average sample depth across interval) are given. How come I often see a combination like this?
TF:IDP NO_READS:96.09
Filtered as NO_READS, yet average depth of 96.09 across the interval? Of course, PART of the interval may be without reads, even more than the threshold set by --coverage_status_threshold, but this is what I understand is meant by COVERAGE_GAPS. I also wondered whether the reads might all have been totally filtered out due to low quality parameters and adjusted --minimum_base_quality and --minimum_mapping_quality to 0, but NO_READS are still flagged for a number of intervals, despite IDP being far from 0. Is this a bug, or have I misunderstood something?
L. Pihlstrom