Hi,

I would like to calculate linkage disequilibrium among MHC alleles in a non-model organism, where a genotype contains multiple alleles found across duplicated loci showing copy number variation. We do not know the locus affiliation of alleles (as we use degenerate primers) and so this renders the typical approach used by LD softwares invalid as alleles need to be assigned to loci (in those that I have come across).

Would it not be valid to calculate all possible pairwise combinations of alleles in a population and test the deviation of the frequency of pairs from expectations under Hardy Weinberg equilibrium? This totally ignores assigning alleles to loci.

Any advice would be much apprecoiated.

Many thanks in advance 

J