changed default AF for Mutect2 tumor-only

[davidbenjamin]

This value is more correct. I didn't use it earlier because it was protecting against some false positives, but those were not contamination or germline (i.e. default AF was filtering them for the wrong reasons) and we can instead filter them by catching variants on the same haplotype as filtered variants.

This improves sensitivity in MC3.

[codecov-io]

Codecov Report

Merging #5067 into master will increase coverage by 0.022%.
The diff coverage is n/a.

@@              Coverage Diff               @@
##             master     #5067       +/-   ##
==============================================
+ Coverage     86.36%   86.382%   +0.022%     
- Complexity    28830     28881       +51     
==============================================
  Files          1791      1791               
  Lines        133610    133773      +163     
  Branches      14921     14942       +21     
==============================================
+ Hits         115386    115556      +170     
+ Misses        12820     12812        -8     
- Partials       5404      5405        +1

Impacted Files	Coverage Δ	Complexity Δ
...der/tools/walkers/mutect/M2ArgumentCollection.java	93.333% <ø> (ø)	3 <0> (ø)	⬇️
...walkers/bqsr/AnalyzeCovariatesIntegrationTest.java	93.902% <0%> (+0.056%)	40% <0%> (+16%)	⬆️
...e/hellbender/engine/spark/SparkContextFactory.java	73.973% <0%> (+2.74%)	11% <0%> (ø)	⬇️
...ls/genomicsdb/GenomicsDBImportIntegrationTest.java	93.273% <0%> (+3.916%)	108% <0%> (+35%)	⬆️