What are considered valid sample names for PURPLE?

hartwigmedical / hmftools

Various algorithms for analysing genomics data

GNU General Public License v3.0

181 stars 56 forks source link

What are considered valid sample names for PURPLE? #386

Closed toddajohnson closed 1 year ago

toddajohnson commented 1 year ago

I was last successfully using PURPLE 3.8 and just tried 3.8.2, but PURPLE now is failing with "ERROR ... has invalid sample names" 14:23:49.303 [ERROR] vcf(DIR_PATH/SAGE-3.2/germline_genomewide/PAVE-1.4.1/RK001.germline.sage.PAVE.filtered.vcf.gz) has invalid sample names: [RK001_R, RK001_T] 14:23:49.541 [ERROR] vcf(DIR_PATH/SAGE-3.2/somatic/PAVE-1.4.1/RK001.somatic.sage.filtered.vcf.gz) has invalid sample names: [RK001_T, RK001_R]

What are "valid" sample names?

charlesshale commented 1 year ago

Purple now requires that:

for somatic VCFs (Pave somatic and Gripss somatic), the sample name passed in the -tumor config is the second sample in the VCF, and the the sample name passed in the -reference config is the first sample in the VCF
for germline VCFs (Pave germline and Gripss germline), the reverse is true - ie the tumor sample is first in the VCF and the reference is second

The HMF pipeline ensure that this is the case when Pave and Gripss somatic and germline processes are run.

toddajohnson commented 1 year ago

Thanks for the answer. I just realized one step in my pipeline was switching the order of the tumor and reference, but until now, it did not seem to make a difference.