Calculating Tajima's D from SNP frequency data

Question

[mphillips6789]

I have recently run some forward simulations that produce a table with with the major allele frequency for every SNP along single chromosome in my simulations populations. The table looks something like:

chr pos ref pop1    pop2    pop3
3R  56  A   0.833   0.902   0.523
3R  89  C   0.711   0.622   0.623
...

I was primarily interested in doing some heterzygosity calculations and Fst estimations to compare to results from my real data set, which was straight forward enough. However, if possible, I would also like to do Tajima's D calculations across 50kb windows from this simulated data set. Apologize if this is actually something really simple, but not quite sure how to do this with just SNP frequencies based on what I have read on how to do this calculation.

[Kevin Blighe]

Did you ever find a way to do this?

[Jautis]

I don't think it's possible from this summarized data. Tajima's D is calculated within a population and requires knowing the number of pairwise differentiated sites between individuals.