VCFtools filtration strategy --hwe

Question

[chenyangkang24]

Hi everyone,

I'm currently confused by the option of VCFtools: --hwe. From my understanding, this option measure whether each site from every individual in the VCF file significantly departures from the HW balance. But what if individuals in the file are not from the same population or even species? They definitely cannot reach HW balance for they breed separately. In this case what's the meaning of it?

Thanks!

Answer 1

[pufferfish]

You are correct. This option calculates the HWE score and then filters sites according to the threshold you provide.

It would not be a valid filter if any of the following:

• different species (??)
• different populations
• recently admixed populations
• a variant under strong selection

This is a common challenge in bioinformatics: just because an option exists doesn't mean it applies to your particular dataset.

[chenyangkang24]

Thank you,

That's exactly what I thought. I asked this "stupid" question because some tutorial add this option without explanation, even if the samples comes from different species. In this case many SNPs were gone.

Yangkang