Nanopolish fails to call variant when sample is compound heterozygote in that position

[marcotoffoli]

Dear Jared,

Thank you again for the great software. I am having some issues with one variant not being called correctly from Nanopolish. In that position, the reference has a T, while my sample has C(55%), G(28%) and T(12%). I've done some research and it looks like the reference I'm using (hg19) has a mistake in this position, as the reported frequencies in the population are C(50%) and G(50%), with T at 1-2%.

My assumption is that the correct genotype is C/G, but Nanopolish is calling it T/C heterozygote. The support fractions with the --calculate-all-support enabled are 0.020,0.530,0.426,0.025

I am interested in hearing what you think about this.

[jts]

Hi @marcotoffoli,

Yes, this is a limitation in the way nanopolish internally stores variants that I recently discovered after a discussion with another user. It will be somewhat difficult to fix but I'll leave this issue open as a reminder.

Jared

[marcotoffoli]

Thank you! For the time being, I just edited the reference and after that Nanopolish correctly calls the variant.

Cheers Marco

[jts]

Ok, I'm glad it is working after editing the reference.