Is nucleotide diversity higher than expected by chance?

[JackieBadger]

Hi,

I want to test if multiple alleles in individuals are more similar to each other (in nucleotide diversity) than expected by chance. Does anyone know a way to calculate this over 100's or 1000's of samples?

Cheers,

j

[syfo]

I'd say the problem is to define what can be expected "by chance". Theoritical models usually rely on some unrealistic assumptions (a uniform base distribution along the genome for instance). Do you have a negative control, like a set of other alleles (or genomic positions) you do not expect to be specially constrained? You could compare the variability in this group with the one in your alleles of interest.
If the only sequence information you have for each sample is limited to your alleles of interest then I have no idea, unless you can distinguish subgroups of alleles according to some relevant criteria. Not sure I understood your problem though.