Add HG002 ONT SV genotyping test data

[aksenia]

Adds a small HG002 ONT subset (BAM 0.77 MB + index) and a matching VCF of 15 known structural variants (INS, DEL, INV, BND) for testing SV genotyping tools that genotype a given SV set, such as Sniffles --genotype-vcf. They are needed by the upcoming sniffles/genotypevcf module in nf-core/modules.

The BAM covers four small windows of GRCh38, subsampled to ~12x by read name, with base qualities, MM/ML/mv/ts/ns/HP/PS tags, @PG lines and UR paths removed. The README in the folder documents provenance, files and the expected genotypes (1 0/0, 7 0/1, 7 1/1 with Sniffles 2.8.1).

Files: data/genomics/homo_sapiens/nanopore/bam/HG002_ont_sv_genotype/

No existing long-read BAM in test-datasets has a matching known-SV VCF at a depth where heterozygous and homozygous genotypes can be called. The SAVANA COLO829 subset (~1-2x, no SV set) and the generic nanopore/bam/test.sorted.bam (contigs < 1 Mb, which Sniffles skips) were considered.

[mashehu]

would it be possible to make it <1MB or even smaller? the bigger the test-dataset the longer tests take (and the more bloated the repo gets) 🙂

[aksenia]

thanks for having a look @mashehu ! Yes I can reduce number of regions and re-sample. I actually looked for nf-core modules guideline about the advised size limits and did not find any. Did I miss it? Patch incoming.

[mashehu]

we are rather vague by saying "as small as possible". but you can look at other data sets where we for example often just subsample to chr22

[aksenia]

re-uploaded smaller dataset @mashehu.

[mashehu]

please drop the old file from the previous commit and force push, to keep the git history clean 🙂

[aksenia]

OK done @mashehu :)

[mashehu]

Great, thanks!