作者: Stanley Sawyer , Daniel L. Hartl , Carlos D. Bustamante , John Wakeley
DOI: 10.1093/GENETICS/159.4.1779
关键词: Biology 、 Directional selection 、 Random field 、 Negative selection 、 Population genetics 、 Poisson distribution 、 Covariance 、 Statistics 、 Confidence interval 、 Genetics 、 Robustness (evolution)
摘要: In this article we explore statistical properties of the maximum-likelihood estimates (MLEs) selection and mutation parameters in a Poisson random field population genetics model directional at DNA sites. We derive asymptotic variances covariance MLEs power likelihood ratio tests (LRT) neutrality for varying levels as well robustness LRT to deviations from assumption free recombination among also discuss coverage confidence intervals on basis two standard-likelihood methods. find that has high detect estimation performs very when ancestral states all mutations sample are known. When not known, test negative but positive selection. is robust independence