Silent "failure" on sex chromosomes?

[jiseonmin]

Hi colab,

First, thank you so much for creating and sharing an awesome tool! I can confirm it was pretty straightforward to install (using pixi) and run smoothly (and so fast..) on my university cluster . But I wanted to flag one thing about ploidy. I was able to run genome_scan.py on X chromosome on a vcf file with just male flies, and it ran fully without a warning. However, the output looks weird since it assumes everyone is homozygous everywhere. Then I saw line 482 in pg_gpu/haplotype_matrix.py (assert ploidy == 2) inside from_vcf, which doesn't seem right, at least without a warning. So my question is, could you edit the line 480 (genotypes = allel.GenotypeArray(vcf['calldata/GT'])) and beyond to correct for pseudo-diploids or print warning and stop if there is mixture of haploids and diploids? And I guess if the region includes both autosomes and sex chromosome, it might be a little more work.
Lastly, I've only used vcf input, so it might be worth checking if similar things happen to other input formats that doesn't depend on from_vcf.

Thanks,
Jiseon

[nspope]

Good point---thanks @jiseonmin

[andrewkern]

yes thanks @jiseonmin! I'll file an issue to add this as a feature and reference this one.