Format

Send to

Choose Destination
Evol Bioinform Online. 2007 Feb 4;2:145-55.

Assessing the applicability of the GTR nucleotide substitution model through simulations.

Author information

1
Laboratory of Evolutionary Genetics, Institute for Molecular Biology and Medicine, Université Libre de Bruxelles, CP300, rue Jeener et Brachet 12, 6041 Gosselies, Belgium.

Abstract

The General Time Reversible (GTR) model of nucleotide substitution is at the core of many distance-based and character-based phylogeny inference methods. The procedure described by Waddell and Steel (1997), for estimating distances and instantaneous substitution rate matrices, R, under the GTR model, is known to be inapplicable under some conditions, ie, it leads to the inapplicability of the GTR model. Here, we simulate the evolution of DNA sequences along 12 trees characterized by different combinations of tree length, (non-)homogeneity of the substitution rate matrix R, and sequence length. We then evaluate both the frequency of the GTR model inapplicability for estimating distances and the accuracy of inferred alignments. Our results indicate that, inapplicability of the Waddel and Steel's procedure can be considered a real practical issue, and illustrate that the probability of this inapplicability is a function of substitution rates and sequence length.We also discuss the implications of our results on the current implementations of maximum likelihood and Bayesian methods.

KEYWORDS:

GTR model; homogeneity; nucleotide substitution; phylogeny inference; simulations

PMID:
19455208
PMCID:
PMC2674669

Supplemental Content

Full text links

Icon for PubMed Central
Loading ...
Support Center