GitHub
09/11/2026, 12:56 PMgnomAD.r2.1.1-sv.vcf.gz used bare chromosome names (22) while sv_query.vcf.gz (the query VCF used against it in `nf-core/modules`' svdb/query test) uses chr-prefixed names (chr22). This mismatch means the two files describe the same genomic positions but never lexically match.
svdb <= 2.8.4 silently stripped chr prefixes on read, so the mismatch didn't matter in practice. svdb 2.9+ intentionally removed that normalization, so with svdb 2.12.0 the two files no longer overlap at all, and the svdb/query test's only annotated hit silently disappeared.
This PR fixes that by adding chr prefix to `gnomAD.r2.1.1-sv.vcf.gz`'s CHROM column, ##contig header IDs, and CHR2 INFO values so it's consistent with sv_query.vcf.gz, restoring the intended annotation match under current and future svdb versions.
This file is only referenced by `nf-core/modules`' svdb/query test, so this change is isolated.
nf-core/test-datasetsGitHub
09/11/2026, 1:42 PMGitHub
09/11/2026, 1:43 PM