PGDSPider VCF to BAYENV

Question

[biosol]

Hi all,

I'm trying to convert my VCF file to Bayenv2 format with PGDSpider 2.1.1.5. My problem is that it recognises a lot of SNPs as non-polymorphic (although they actually are...). I've realised that some of my SNPs have in the VCF REF column more than just one alelle (I guess they are insertions) and it's only those SNPs the ones that PGDSpider recognises as polymorphic.

I don't know if someone is facing the same problem, but I'd be glad to discuss what can be done, because I'd like to keep all those SNPs in my analysis. Thank you very much in advanced!

[brallen]

Hi, Sonia.

I see you posted this question over a year ago. Did you find a resolution? I'm having the same issue. During the conversion from VCF to Bayenv format, all SNPs are removed because they aren't polymorphic (when in reality, they are). I'd appreciate any advice!

Cheers, Brian

[biosol]

Hi Brian! for me it was something super silly: in the population file I had spaces instead of tabs between the columns (or the other way around, I don't remember...)! Maybe it is your problem as well. Also, I remember that I played around with the .spid file. For the VCF to Bayenv conversion there is a question that says: "# Do you want to include non-polymorphic SNPs?" Maybe if you keep having the problem, you can tell it to include all the SNPs (even if it doesn't recognise them as polymorphic). I hope this helps, let me know if I can help you with something else!

[brallen]

Thanks, Sonia! I really appreciate the help. Yes, the issue was definitely the population definition file.. mostly that I didn't include one! Once I added the popdef with the correct ID's, I was able to correctly convert the VCF to bayenv2 format and use it with the bayenv2 software. Interestingly, the standard output file still says the loci were removed for not being polymorphic, but I think this is a bug because the other output file correctly lists the SNP counts for each population. Best, Brian