Make M2 haplotype and clustered events filters smarter about germline events

[davidbenjamin]

Closes #7391

[gatk-bot]

Github actions tests reported job failures from actions build 8172680846
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
variantcalling	17.0.6+10	8172680846.2	logs

[gatk-bot]

Github actions tests reported job failures from actions build 8181065857
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
variantcalling	17.0.6+10	8181065857.2	logs
variantcalling	17.0.6+10	8181065857.2	logs

[codecov]

Codecov Report

Merging #8717 (be4b7de) into master (c97faf6) will decrease coverage by 2.767%.
The diff coverage is 87.500%.

❗ Current head be4b7de differs from pull request most recent head 2f1c832. Consider uploading reports for the commit 2f1c832 to get more accurate results

Additional details and impacted files

@@               Coverage Diff               @@
##              master     #8717       +/-   ##
===============================================
- Coverage     86.407%   83.640%   -2.767%     
+ Complexity     39882     38213     -1669     
===============================================
  Files           2371      2371               
  Lines         187596    187624       +28     
  Branches       20532     20537        +5     
===============================================
- Hits          162096    156928     -5168     
- Misses         18272     23475     +5203     
+ Partials        7228      7221        -7     

Files	Coverage Δ
...alkers/mutect/filtering/ClusteredEventsFilter.java	100.000% <100.000%> (ø)
...kers/mutect/filtering/FilteredHaplotypeFilter.java	91.111% <100.000%> (-6.667%)	⬇️
...e/hellbender/utils/variant/GATKVCFHeaderLines.java	95.522% <100.000%> (ø)
...r/tools/walkers/mutect/Mutect2IntegrationTest.java	5.405% <0.000%> (-89.189%)	⬇️
.../tools/walkers/mutect/SomaticGenotypingEngine.java	83.258% <88.571%> (-10.524%)	⬇️

... and 141 files with indirect coverage changes

[ldgauthier]

How hard would it be to put in a regression test where you do something similar to #7391 and compare with and without --genotype-germline-sites? There's not a lot of coverage for that arg as-is. And I'm a little concerned that while the ECNT annotation changed in the SW test, there were no filtering changes in order tests.

[davidbenjamin]

Added a regression test with two good variants that this PR rescues from being false negatives. Also made the sensitivity thresholds in the other integration tests higher because even without --genotype-germline-sites this PR saves a few calls.

[gatk-bot]

Github actions tests reported job failures from actions build 8349590850
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
variantcalling	17.0.6+10	8349590850.2	logs

[ldgauthier]

I saw your "whoops" commit -- I have actually merged commits that turn off tests. James is trying to figure out if we can have codecov check the absolute number of tests since that should almost never go down. He claims you can run select tests from the data provider with IntelliJ, but it seems... cumbersome.

[ldgauthier]

Thanks for the new test! The sensitivity threshold is a good find too

[gatk-bot]

Github actions tests reported job failures from actions build 8354454301
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
variantcalling	17.0.6+10	8354454301.2	logs

[gatk-bot]

Github actions tests reported job failures from actions build 8384399332
Failures in the following jobs:

Test Type	JDK	Job ID	Logs
variantcalling	17.0.6+10	8384399332.2	logs