Are these filter out sites not used in the calculation?

[jiangzy26]

Hi,
I know we need a hard filter to variant and invariant data, respectively, which means we may lose sites within each sliding window, for example, for 50k window size, after filtering out the low QUAL sites, we may only have 40k sites, I am not sure whether we need to add invariants to these 'lost' positions, could you share your help? Thanks.

[hrluo93]

Hi,

I am also want to know how to covert low quality variant site to invariant site. I see a pipeline seems can keep invariant sites but filtered low quality variant sites, and combine invariant site+high quality variant site. I don't know if it suitable.

you can check the website:
https://yanzhongsino.github.io/2023/03/13/bioinfo_population.genetics_pixy/

Best wishes!