ways to verify positive selection of a mutation with vcf data

Question

[honglinzhou]

Hi every one,

We sequenced the genomes for about 20 patients having a disease from a family and and verified all of them have the disease-causing snp. Now we got another idea that the disease could just be a byproduct and this mutant could be favored (positive selected) in the population due to other possible benefits. So we like to know whether there is positive selection with the disease-causing snp in the population. Can we test it and how to do it? we also have hundreds control genomes too.

Somebody suggested phasing these genomes and comparing the hyplotype diversity between the patient group and controls. But I am not sure whether it works.

Comments for both the idea and methods are very welcome!

Thanks

Robin