Some general likelihood and Bayesian methods for analyzing single nucleotid
e polymorphisms (SNPs) are presented. First, an efficient method for estima
ting demographic parameters from SNPs in linkage equilibrium is derived. Th
e method is applied in the estimation of growth rates of a human population
based on 37 SNP loci. It is demonstrated how ascertainment biases, due to
biased sampling of loci, can be avoided, at least in some cases, by appropr
iate conditioning: when calculating the likelihood function. Second, a Mark
ov chain Monte Carl (MCMC) method for analyzing linked SNPs is del eloped.
This method call be used for Bayesian and likelihood inference on linked SN
Ps. The utility of the method is illustrated by estimating recombination ra
tes in a human data set containing 17 SNPs and 60 individuals. Both methods
are based on assumptions of low mutation rates.