Topological Bias in Distance-Based Phylogenetic Methods: Problems with Over- and Underestimated Genetic Distances

摘要: I show several types of topological biases in distance-based methods that use the least-squares method to evaluate branch lengths and minimum evolution (ME) or Fitch-Margoliash (FM) criterion choose best tree. For a 6-species tree, there are two tree shapes, one with three cherries (a cherry is pair adjacent leaves descending from most recent common ancestor), other two. When genetic distances underestimated, 3-cherry shape favored either ME FM criterion. overestimated, favors 2-cherry but direction bias depends on whether negative branches allowed, i.e. allowing disallowing shape. The extent explored by computer simulation sequence evolution.