search

dnanexus-rnd / GLnexus

Scalable gVCF merging and joint variant calling for population sequencing projects
Apache License 2.0
145 stars 37 forks source link

Use min_AQ1 correctly when min_AQ2=0 #167

Closed mlin closed 5 years ago

mlin commented 5 years ago

Follows up #164 with various adjustments and assertions

mlin commented 5 years ago

@tedyun There were some downstream effects of the change to initializing top_AQ::V to -1 instead of 0, but I think I plowed through them here, and that this is a good solution to the problem.

One question, do you think the min_AQ1>0,min_AQ2=0 condition is one you're likely to use in practice? If so, then we should add a specific test case for it. But if it's more of a dusty corner that you stumbled on and aren't likely to revisit, then I'm less worried about that.

tedyun commented 5 years ago

@mlin Thank you very much Mike!

No, there is little chance we will use min_AQ1>0, minAQ2=0 in practice. We were experimenting with a wide spectrum of AQs and we've stumbled upon it :)