摘要：The software POPS performs inference of population genetic structure using multilocus genotypic data. Based on a hierarchical Bayesian framework for latent regression models, POPS implements algorithms that improve estimation of individual admixture proportions and cluster membership probabilities by using geographic and environmental information. In addition, POPS defines ancestry distribution models allowing its users to forecast admixture proportion and cluster membership geographic variation under changing environmental conditions. We illustrate a typical use of POPS using data for an alpine plant species, for which POPS predicts changes in spatial population structure assuming a particular scenario of climate change.
关键词：latent class regression models;mixture models;MCMC;population genetic structure;environmental covariates
