VCF to R problems

Question

[BeetleCheers]

Hi there,

I am having trouble with my vcf files in R. These files were generated by custom perl wrappers that automate STACKS. For context, my study is a landscape pop-gen analysis of pest insects. At first, I could not load the finished files into R at all (30 000 loci unfiltered). I reduced the file down to 10 individuals (from >100), and the file will read into R now using the command:

>MPB10vcf=read.vcf('10_thinned.recode.vcf')

However, I can't seem to do anything with it now that it's there.

I would like to convert the vcf data into a genind object so I can run a PCA/DAPC and graph out the genetic structure. However, I am getting some different errors each time try to do this:

> MPB10_genind=vcfR2genind(MPB10vcf)
#Error in extract.gt(x) : ID column contains non-unique names

Does anyone have any thoughts/experiences with this?

Answer 1

[Philipp Bayer]

I think this happened, here's an example row from my own data which triggers the same error:

Chrom1    3653    .    A       G       86      PASS    NS=2    GT:DP   1/1:39  0/0:47  0/0:5   0/0:5   1/1:5   1/1:5   0/0:5   1/1:5   1/1:5   0/1:5   1/1:5   1/1:5   1/1:5   0/0:5   0/0:5   0/1:5   0/0:5   0/0:5   0/0:5   0/0:5   0/0:5

The third column is the ID column, which in my case is a dot for all SNPs! I have to replace these by something unique (Chrom+Position should work), then vcfR2genind should work too.

By the way, vcfR comes with its own vcf loading command, no need to use pegas' read.vcf: read.vcfR()

[BeetleCheers]

Thanks! I can load the modified file into R now, though I think there are still some troubles with how the original VCF file was generated. Ever forward!