Performance-Based Selection of Likelihood Models for Phylogeny Estimation

Abstract

Phylogenetic estimation has largely come to rely on explicitly model-based methods. This approach requires that a model be chosen and that that choice be justified. To date, justification has largely been accomplished through use of likelihood-ratio tests (LRTs) to assess the relative fit of a nested series of reversible models. While this approach certainly represents an important advance over arbitrary model selection, the best fit of a series of models may not always provide the most reliable phylogenetic estimates for finite real data sets, where all available models are surely incorrect. Here, we develop a novel approach to model selection, which is based on the Bayesian information criterion, but incorporates relative branch-length error as a performance measure in a decision theory (DT) framework. This DT method includes a penalty for overfitting, is applicable prior to running extensive analyses, and simultaneously compares all models being considered and thus does not rely on a series of pairwise comparisons of models to traverse model space. We evaluate this method by examining four real data sets and by using those data sets to define simulation conditions. In the real data sets, the DT method selects the same or simpler models than conventional LRTs. In order to lend generality to the simulations, codon-based models (with parameters estimated from the real data sets) were used to generate simulated data sets, which are therefore more complex than any of the models we evaluate. On average, the DT method selects models that are simpler than those chosen by conventional LRTs. Nevertheless, these simpler models provide estimates of branch lengths that are more accurate both in terms of relative error and absolute error than those derived using the more complex (yet still wrong) models chosen by conventional LRTs. This method is available in a program called DT-ModSel.

Nitrogen fixation in a chemoautotrophic lucinid symbiosis

“…ProtTest version 3 (refs 78,79) was employed to select the best-fit model of aminoacid replacement from 120 models (15 matrices: +G, +I or + G+I; +F) with a starting topology based on maximum likelihood. As the Bayesian information criterion (BIC; also known as the Schwarz criterion, or SC) 80 , the corrected Akaike information criterion (AICc) 81,82 and the decision theory framework (DT) 83 favoured LG+I+G, the phylogeny was then inferred using RAxML version 8.2.4 (ref. 84) with the LG amino acid matrix 85 , a gamma model of rate heterogeneity and an estimate of proportion of invariable sites.…”

Phylogenetic analysis of the insulin-like growth factor binding protein (IGFBP) and IGFBP-related protein gene families

“…Ten million generations were performed with four chains (Markov Chain Monte Carlo) and a tree was saved every 100 generations. Priors included a mixed amino acid model allowing for optimization of the model during the analysis (Minin et al, 2003). Multiple analyses were started from different random locations within the tree space in order to test for the occurrence of stationarity, convergence and mixing within the ten million generations.…”

Identifying cryptic diversity with predictive phylogeography

“…We used BEAST v. 1.8.2 [51] to infer topology and divergence times with mitochondrial or chloroplast sequence data for SAL and PNW taxa. A model of sequence evolution for each taxon was selected using DT-ModSel [52] (electronic supplementary material, table S5 and see electronic supplementary material, S1 for more details). The Markov chain Monte Carlo analysis was run for 100 million generations, sampling every 1000 generations, with a random starting tree and strict molecular clock.…”