Power analysis for SNP data

Question

[cecilia.villacorta]

What is the best tool for Power Analysis of SNP data? I have tried POWSIM, but it has a limit of 50 loci (Ryman & Palm 2006 - Molecular Ecology Notes). Can somebody recommend another way of performing the analysis?

Thanks!

[karl.stamm]

Forget the SNP data; the genotype is just a category. Look for power analyses in general. What kind of hypothesis are you trying to do? Group-quantitative analyses could use a two sample t-test, which is much simpler to estimate power over.

[cecilia.villacorta]

Thanks Karl. If I am not wrong, POWSIM actually does a chi-square test. I want to test different datasets and see what is their power for estimating Fst values.

[karl.stamm]

Okay, thanks for clarifying. A chi-square test is about likelihoods of genotypes, to see if a sample is evenly distributed or selected. When you say power for estimating Fst values, that's something else. I think you mean precision and accuracy. The calculation of Fst is not straightforward like a chi-square. You should find a formula for Fst and see where parameters like subject count and genotype frequency and density fit. I don't know any tools to help; you should re-title the question to ask for "precision of Fst scores" and see if anyone knows a tool to calculate that.