fixed incoherent unit test cases in allele subsetting utils #6448
Conversation
...ava/org/broadinstitute/hellbender/tools/walkers/genotyper/AlleleSubsettingUtilsUnitTest.java
Outdated
Show resolved
Hide resolved
...ava/org/broadinstitute/hellbender/tools/walkers/genotyper/AlleleSubsettingUtilsUnitTest.java
Show resolved
Hide resolved
...ava/org/broadinstitute/hellbender/tools/walkers/genotyper/AlleleSubsettingUtilsUnitTest.java
Outdated
Show resolved
Hide resolved
...ava/org/broadinstitute/hellbender/tools/walkers/genotyper/AlleleSubsettingUtilsUnitTest.java
Outdated
Show resolved
Hide resolved
| attribute(GATKVCFConstants.STRAND_COUNT_BY_SAMPLE_KEY, new int[]{10, 10, 11, 11}).GQ(200).make())}); | ||
|
|
||
| tests.add(new Object[]{ | ||
| new VariantContextBuilder(vcBase).alleles(ACG).genotypes(new GenotypeBuilder(base).alleles(AA).AD(hetCG3AllelesAD).PL(hetCG3AllelesPL). |
There was a problem hiding this comment.
What was wrong with this test case?
There was a problem hiding this comment.
Based on the ADs and PLs this case was supposed to be: triallelic (A,C,G) variant with het alt (C/G) genotype, subset to just the ref and one alt allele. But that throws an error since the genotype has both alt alleles. Previously no error was thrown because the genotype was accidentally set as hom ref, contradicting the ADs and PLs. The test as intended was inherently impossible.
There was a problem hiding this comment.
Got it. You can take it out entirely then.
There was a problem hiding this comment.
It's gone; the diff mixed-and-matched similar lines to give the appearance that part of this case remained.
|
Ready for round 2 of review. |
davidbenjamin
force-pushed
the
db_allele_subsettingtest
branch
from
21c2d05
to
306dd06
Compare
davidbenjamin
force-pushed
the
db_allele_subsettingtest
branch
from
306dd06
to
a4e6390
Compare
Closes #3672.
@ldgauthier The ADs, GTs, and PLs were inconsistent.