vcftools Tajima's D calculation

Question

[biogirl]

Hi all,

I'm running vcftools on a vcf (obv, ha) to calculate Taijma's D in sliding windows of 10,000 bp:

vcftools --vcf in.vcf --out tajimasd --TajimaD 10000

The log file states that after filtering, kept 342087 out of a possible 342087 Sites (so all sites). Yet when I look at the output file, there are a lot of 'nan' (meaning no SNPs in that bin), and the bins that have Tajima's D calculated have a tiny number of SNPs in them (max 8 - certainly not adding up to >300k SNPs).

Any ideas as to why I'm getting no SNPs in the majority of my bins?

Thanks

Answer 1

[grey]

I believe this is because vcftools --TajimaD doesn't like missing genotype calls ./. and will not calculate Tajimas D there

See: vcftools --TajimaD does not start at 0 and misses variants