Validating HG002

[skoren]

I'm trying to run veritymap on an HG002 assembly with HiFi data (4 cells, about 35x). The problem is I end up with a very tiny sam file and most regions have no reads aligned when I view the HTML and the following coverage histogram:

The stats for norarekmers.bed reports more than 1/3 of the assembly length has no markers. I've successfully run veritymap on other genomes and more diverse diploid genomes and get reasonable coverage plots:

Is there anything I can do to improve the mappings here or is veritymap not suitable for diploid human genomes yet? Do I need higher HiFi coverage?

[seryrzu]

Unfortunately, VerityMap doesn't work with diploid homogeneous genomes since most k-mers are not specific to a particular contig (they are shared b/w homologs). We are working on going around this issue. Sorry for the inconvenience.