Likelihood ratio tests LRTs based on codon models of evolution.
收藏Figshare2015-12-02 更新2026-04-29 收录
下载链接:
https://figshare.com/articles/dataset/_Likelihood_ratio_tests_LRTs_based_on_codon_models_of_evolution_/424016
下载链接
链接失效反馈官方服务:
资源简介:
M3* is the HYPHY implementation of model M3 [98]. Tests comparing M3* vs. Dual and M7 vs. M8 were performed only for genes where M0 vs. M3 was significant. The test FMutSel0 vs. FMutSel was performed only for genes with large samples.aResults for lb genes are shown for full samples. However, all LRT results remain the same when only D. melanogaster was analyzed (with or without the recombinants).bLRT statistic is double the difference between the likelihood values optimized under the alternative and null models: 2(???alt – ???null). P-value is computed using the χ2-distribution with d.f. = 4 for LRTs of M0 vs. M3, and M3* vs. Dual. For LRT of M7 vs. M8 d.f. = 2, and for FMutSel0 vs. FMutSel d.f. = 41. For the LRT of positive selection on noncoding regions d.f. = 1. Significant P-values in bold.cEstimates under M0: ω = dN/dS; tree length t is measured by number of expected nucleotide substitutions per codon over the tree; transition/transversion ratio κ (estimated under M3, or under M0 if the test M0 vs. M3 is not significant); dS, tree length for synonymous substitutions , dN, tree length for nonsynonymous sites. Estimates under Dual model: coefficients of variation ( = standard deviation/mean) for distributions of synonymous and nonsynonymous rates, CVS and CVN respectively. Estimates under mutation-selection model FMutSel: P+ is the proportion of mutations with advantageous effect (S = 2Ns>0); S+ is mean selection coefficient of all advantageous mutations; S– is mean selection coefficient of all deleterious mutations. Estimates under M8: p, q are parameters controlling the shape of the Beta-distribution; p0 is proportion of sites in the sequence with ω-values from beta-distribution (between 0 and 1); ω = dN/dS, with discrete class allowed to be >1 (under positive selection); p1 is proportion of sites in the discrete class. In the combined coding and noncoding model [102], sites in the noncoding region come from two categories: proportion p0 are under negative selection with ζ0p1 are under positive selection or evolving neutrally with ζ1≥1; ζ0 and ζ1 are estimates of the ratio of substitution rate in the noncoding region to average synonymous rate in the coding region for the two categories.
创建时间:
2015-12-02



