Gene coverage issue in HG38

Question

[rajesh.msch]

Hi, Recently I checked the coverage of gene OCA2, using both genome HG37 and hg38.

surprisingly, it was not not covered (very poorly mapped) in HG38, and full covered with good coverage in hg19 with all used capture kit.

we are not able to identify the problem, because for both of the genome we used GATK 4.3.0.0 with updated BWA-MEM aligner.

Suggestion welcome.

[i.sudbery]

I think there is a typo in your question. You can that it was not covered in HG38, but was covered in hg38. Presumably you mean one of these to be hg37?