Laura Gauthier

@DarioS GATK doesn't have a tool that will add more variant alleles to a sample's genotype. Low AC alleles can get dropped at multi-allelic sites resulting in the called GT...

Hi @elcortegano, A lot of things changed in GATK3 with the advent of GVCFs. We don't store annotation data for reference sites in GVCFs, partly because in most cases sites...

@lbergelson am I reading the spec wrong? I thought `@RG` was required?

Got it. Then I guess I'm arguing against @aushev's expected behavior that > HaplotypeCaller should just treat them normally without any error

I really don't want to assume that the data is correct because in the event that a user does try to run GVCF mode on a multi-sample bam without an...

I opened #6511 for SplitNCigarReads and https://github.com/broadinstitute/picard/issues/1488 for MarkDuplicates. They should both be easy to address once we decide on warning or error.

From discussion over in the Picard repo, I've realized that I read the spec wrong. In that case, I'm willing to let HaplotypeCaller "power through" after emitting a warning if...

I would have guessed that GenotypeGVCFs is reproducing the phasing from the PGT FORMAT tag, but then they don't agree in the output. I trust HaplotypeCaller more. I'll take a...

After Jonathan's recent modifications I'm a little confused about what should happen if a reference isn't provided -- should it ever error or just warn? @jamesemery I ended up changing...

I feel like this should happen all the time when there are phased calls. You're telling me that for ``` GT:AD:DP:GQ:PGT:PID:PL 0/1:44,6:50:99:0|1:14590_G_A:135,0,2509 0/1:28,8:36:99:0|1:14599_T_A:249,0,1923 0/1:21,6:27:99:0|1:14590_G_A:186,0,1528 0/1:42,9:51:99:0|1:14599_T_A:210,0,2353 0/0:11,0:11:0:.:.:0,0,163 0/0:4,0:4:12:.:.:0,12,119 0/0:15,0:15:39:.:.:0,39,585 ./.:0,0:0:.:.:.:0,0,0 0/0:11,0:11:30:.:.:0,30,382...