The Prevalence and Impact of Model Violations in Phylogenetic Analysis

Abstract

In phylogenetic inference, we commonly use models of substitution which assume that sequence evolution is stationary, reversible, and homogeneous (SRH). Although the use of such models is often criticized, the extent of SRH violations and their effects on phylogenetic inference of tree topologies and edge lengths are not well understood. Here, we introduce and apply the maximal matched-pairs tests of homogeneity to assess the scale and impact of SRH model violations on 3,572 partitions from 35 published phylogenetic data sets. We show that roughly one-quarter of all the partitions we analyzed (23.5%) reject the SRH assumptions, and that for 25% of data sets, tree topologies inferred from all partitions differ significantly from topologies inferred using the subset of partitions that do not reject the SRH assumptions. This proportion increases when comparing trees inferred using the subset of partitions that rejects the SRH assumptions, to those inferred from partitions that do not reject the SRH assumptions. These results suggest that the extent and effects of model violation in phylogenetics may be substantial. They highlight the importance of testing for model violations and possibly excluding partitions that violate models prior to tree reconstruction. Our results also suggest that further effort in developing models that do not require SRH assumptions could lead to large improvements in the accuracy of phylogenomic inference. The scripts necessary to perform the analysis are available in https://github.com/roblanf/SRHtests, and the new tests we describe are available as a new option in IQ-TREE (http://www.iqtree.org).

Phylogenomic relationships of bioluminescent elateroids define the ‘lampyroid’ clade with clicking Sinopyrophoridae as its earliest member

“…Heterogeneous partitions were excluded from the dataset A‐4199‐AA to generate dataset F‐2195‐AA. We used Maximum Symmetry Test (Naser‐Khdour et al ., 2019) to exclude the deviating genes from the dataset B‐4199‐NT ( P ‐value cutoff <0.05) and the dataset G‐958‐NT contains only partitions that passed the test. The software SymTest v.2.0.49 (https://github.com/ottmi/symtest) was used to calculate the deviation from stationarity, reversibility and homogeneity (Jermiin et al ., 2008) (SRH).…”

Spaghetti to a Tree: A Robust Phylogeny for Terebelliformia (Annelida) Based on Transcriptomes, Molecular and Morphological Data

“…We identified loci or gene trees with heterogeneity using three metrics. First, we excluded loci that violated the assumptions of substitution models that sequence evolution is stationary, reversible and homogenous, by using the matched-pairs test of symmetry [42] implemented in IQ-TREE v. 2 [43]. Second, we identified gene trees with heterogeneity in branch lengths that may cause long branch attraction, by measuring the standard deviation of long branch scores (LB_SD) [44] for each gene tree.…”

Assessing Confidence in Root Placement on Phylogenies: An Empirical Study Using Non-Reversible Models for Mammals

“…There is strong empirical evidence that molecular evolutionary processes are rarely reversible (Squartini and Arndt 2008;Naser-Khdour, et al 2019), but few studies have explored the accuracy of non-reversible substitution models to root phylogenetic trees (Huelsenbeck, et al 2002;Yap and Speed 2005;Williams, et al 2015;Cherlin, et al 2018;Bettisworth and Stamatakis 2020). Most studies that have looked at this question in the past have focused on either simulated datasets (Huelsenbeck, et al 2002;Jayaswal, et al 2011;Cherlin, et al 2018;Bettisworth and Stamatakis 2020) or relatively small empirical datasets (Yang and Roberts 1995;Yap and Speed 2005;Jayaswal, et al 2011;Heaps, et al 2014;Williams, et al 2015;Cherlin, et al 2018).…”