Escaping from the Felsenstein Zone by Detecting Long Branches in Phylogenetic Data

摘要: Long branches in a true phylogeny tend to disrupt hierarchical character covariation (phylogenetic signal) the distribution of traits among organisms. The distortion structure character-state matrices can lead errors estimation phylogenetic relationships and inconsistency methods inference. Examination trees distorted by long-branch attraction will not reveal identities problematic taxa, part because mask long reducing inferred branch lengths through branching order. Here we present simple method for detection taxa whose placement evolutionary is made difficult effects attraction. an extension tree-independent conceptual framework data exploration (RASA). Taxa that are likely attract revealed leave distinct footprints states these traces be directly observed error RASA regression. Problematic identified using new diagnostic plot called taxon variance plot, which apparent cladistic phenetic variances contributed individual compared. procedure identifying edges employs algorithms solved polynomial time applied morphological, molecular, mixed characters. efficacy demonstrated simulated evolution empirical evidence set recently published sequences. We show accuracy improved detecting combating potentially misleading influences taxa.