ekg/hgsvc

building variation graphs from hgsvc data

hgsvc's Issues

Not all variants end up in HGSVC.haps.vcf.gz

vcfcombine isn't outputting all the variants I would expect. For example, HG00514 has 598 variants in chr21:

bcftools view ./HG00514.haps.vcf.gz -r chr21 -H | grep -v "0|0" | wc -l
598

But not a single one ends up in ./HGSVC.haps.vcf.gz

bcftools view ./HGSVC.haps.vcf.gz -r chr21 -H -s HG00514 | grep -v "0|0" | wc -l
0

Some, like the second one (chr21 5114345), I can understand as they collide with different variants at the exact same positions in other samples. But unique ones (like the first) don't show up either. This seems to happen in all chromosomes after chr10 for HG00514 and chr16 for NA19240.

URLs to hap data can't be found

The URLs to the server appear to have changed, as the ones in the repo return 404 not found when attempting to download the vcf/tbi files.

Is it possible to update these to the newest version, perhaps on a more persistent filesystem?