Uniformization for Sampling Realizations of Markov Processes: Applications to Bayesian Implementations of Codon Substitution Models

Nicolas Rodrigue 1 Herve Philippe 1 Nicolas Lartillot 2
2 MAB - Méthodes et Algorithmes pour la Bioinformatique
LIRMM - Laboratoire d'Informatique de Robotique et de Microélectronique de Montpellier
Abstract : MOTIVATION: Mapping character state changes over phylogenetic trees is central to the study of evolution. However, current probabilistic methods for generating such mappings are ill-suited to certain types of evolutionary models, in particular, the widely used models of codon substitution. RESULTS: We describe a general method, based on a uniformization technique, which can be utilized to generate realizations of a Markovian substitution process conditional on an alignment of character states and a given tree topology. The method is applicable under a wide range of evolutionary models, and to illustrate its usefulness in practice, we embed it within a data augmentation-based Markov chain Monte Carlo sampler, for approximating posterior distributions under previously proposed codon substitution models. The sampler is found to be more effcient than the conventional pruning-based sampler, with decorrelation times between draws from the posterior reduced by a factor of twenty or more. CONTACT: nicolas.rodrigue@umontreal.ca.
Type de document :
Article dans une revue
Bioinformatics, Oxford University Press (OUP), 2007, 0, pp.0. 〈www.lirmm.fr/mab〉
Liste complète des métadonnées

https://hal-lirmm.ccsd.cnrs.fr/lirmm-00193706
Contributeur : Nicolas Lartillot <>
Soumis le : mardi 4 décembre 2007 - 14:21:04
Dernière modification le : jeudi 24 mai 2018 - 15:59:22

Identifiants

  • HAL Id : lirmm-00193706, version 1

Collections

Citation

Nicolas Rodrigue, Herve Philippe, Nicolas Lartillot. Uniformization for Sampling Realizations of Markov Processes: Applications to Bayesian Implementations of Codon Substitution Models. Bioinformatics, Oxford University Press (OUP), 2007, 0, pp.0. 〈www.lirmm.fr/mab〉. 〈lirmm-00193706〉

Partager

Métriques

Consultations de la notice

103