@ARTICLE{TreeBASE2Ref16246,
author = {Clemens Lakner and P. v. d. Mark and John P. Huelsenbeck and Bret Larget and Fredrik Ronquist},
title = {Efficiency of Markov Chain Monte Carlo Tree Proposals in Bayesian Phylogenetics.},
year = {2008},
keywords = {},
doi = {10.1080/10635150801886156},
url = {},
pmid = {},
journal = {Systematic Biology},
volume = {57},
number = {1},
pages = {86103},
abstract = {The main limiting factor in Bayesian MCMC analysis of phylogeny is typically the efficiency with which topology proposals sample tree space. Here we evaluate the performance of seven different proposal mechanisms, including most of those used in current Bayesian phylogenetics software. We sampled 12 empirical DNA data sets  ranging in size from 27 to 71 taxa and from 378 to 2,520 sites  under difficult conditions: short runs, no Metropoliscoupling, and an oversimplified substitution model producing difficult tree spaces (Jukes Cantor with equal site rates). Convergence was assessed by comparison to reference samples obtained from multiple Metropoliscoupled runs. We find that proposals producing topology changes as a side effect of branch length changes (LOCAL and Continuous Change) consistently perform worse than those involving stochastic branchrearrangements (nearest neighbor interchange, subtree pruning and regrafting, tree bisection and reconnection, or subtree swapping). Among the latter, moves that use an extension mechanism to mix local with more distant rearrangements show better overall performance than those involving only local or only random rearrangements. Moves with only local rearrangements tend to mix well but have long burnin periods, while moves with random rearrangements often show the reverse pattern. Combinations of moves tend to perform better than single moves. The time to convergence can be shortened considerably by starting with a good tree but this comes at the cost of compromising convergence diagnostics based on overdispersed starting points. Our results have important implications for developers of Bayesian MCMC implementations and for the large group of users of Bayesian phylogenetics software.}
}
Citation for Study 1915
Citation title:
"Efficiency of Markov Chain Monte Carlo Tree Proposals in Bayesian Phylogenetics.".
This study was previously identified under the legacy study ID S1892
(Status: Published).
Citation
Lakner C., Mark P., Huelsenbeck J., Larget B., & Ronquist F. 2008. Efficiency of Markov Chain Monte Carlo Tree Proposals in Bayesian Phylogenetics. Systematic Biology, 57(1): 86103.
Authors

Lakner C.

Mark P.

Huelsenbeck J.

Larget B.

Ronquist F.
Abstract
The main limiting factor in Bayesian MCMC analysis of phylogeny is typically the efficiency with which topology proposals sample tree space. Here we evaluate the performance of seven different proposal mechanisms, including most of those used in current Bayesian phylogenetics software. We sampled 12 empirical DNA data sets  ranging in size from 27 to 71 taxa and from 378 to 2,520 sites  under difficult conditions: short runs, no Metropoliscoupling, and an oversimplified substitution model producing difficult tree spaces (Jukes Cantor with equal site rates). Convergence was assessed by comparison to reference samples obtained from multiple Metropoliscoupled runs. We find that proposals producing topology changes as a side effect of branch length changes (LOCAL and Continuous Change) consistently perform worse than those involving stochastic branchrearrangements (nearest neighbor interchange, subtree pruning and regrafting, tree bisection and reconnection, or subtree swapping). Among the latter, moves that use an extension mechanism to mix local with more distant rearrangements show better overall performance than those involving only local or only random rearrangements. Moves with only local rearrangements tend to mix well but have long burnin periods, while moves with random rearrangements often show the reverse pattern. Combinations of moves tend to perform better than single moves. The time to convergence can be shortened considerably by starting with a good tree but this comes at the cost of compromising convergence diagnostics based on overdispersed starting points. Our results have important implications for developers of Bayesian MCMC implementations and for the large group of users of Bayesian phylogenetics software.
External links
About this resource
 Canonical resource URI:
http://purl.org/phylo/treebase/phylows/study/TB2:S1915
 Other versions:
Nexus
NeXML
 Show BibTeX reference
@ARTICLE{TreeBASE2Ref16246,
author = {Clemens Lakner and P. v. d. Mark and John P. Huelsenbeck and Bret Larget and Fredrik Ronquist},
title = {Efficiency of Markov Chain Monte Carlo Tree Proposals in Bayesian Phylogenetics.},
year = {2008},
keywords = {},
doi = {10.1080/10635150801886156},
url = {},
pmid = {},
journal = {Systematic Biology},
volume = {57},
number = {1},
pages = {86103},
abstract = {The main limiting factor in Bayesian MCMC analysis of phylogeny is typically the efficiency with which topology proposals sample tree space. Here we evaluate the performance of seven different proposal mechanisms, including most of those used in current Bayesian phylogenetics software. We sampled 12 empirical DNA data sets  ranging in size from 27 to 71 taxa and from 378 to 2,520 sites  under difficult conditions: short runs, no Metropoliscoupling, and an oversimplified substitution model producing difficult tree spaces (Jukes Cantor with equal site rates). Convergence was assessed by comparison to reference samples obtained from multiple Metropoliscoupled runs. We find that proposals producing topology changes as a side effect of branch length changes (LOCAL and Continuous Change) consistently perform worse than those involving stochastic branchrearrangements (nearest neighbor interchange, subtree pruning and regrafting, tree bisection and reconnection, or subtree swapping). Among the latter, moves that use an extension mechanism to mix local with more distant rearrangements show better overall performance than those involving only local or only random rearrangements. Moves with only local rearrangements tend to mix well but have long burnin periods, while moves with random rearrangements often show the reverse pattern. Combinations of moves tend to perform better than single moves. The time to convergence can be shortened considerably by starting with a good tree but this comes at the cost of compromising convergence diagnostics based on overdispersed starting points. Our results have important implications for developers of Bayesian MCMC implementations and for the large group of users of Bayesian phylogenetics software.}
}
 Show RIS reference
TY  JOUR
ID  16246
AU  Lakner,Clemens
AU  Mark,P. v. d.
AU  Huelsenbeck,John P.
AU  Larget,Bret
AU  Ronquist,Fredrik
T1  Efficiency of Markov Chain Monte Carlo Tree Proposals in Bayesian Phylogenetics.
PY  2008
KW 
UR  http://dx.doi.org/10.1080/10635150801886156
N2  The main limiting factor in Bayesian MCMC analysis of phylogeny is typically the efficiency with which topology proposals sample tree space. Here we evaluate the performance of seven different proposal mechanisms, including most of those used in current Bayesian phylogenetics software. We sampled 12 empirical DNA data sets  ranging in size from 27 to 71 taxa and from 378 to 2,520 sites  under difficult conditions: short runs, no Metropoliscoupling, and an oversimplified substitution model producing difficult tree spaces (Jukes Cantor with equal site rates). Convergence was assessed by comparison to reference samples obtained from multiple Metropoliscoupled runs. We find that proposals producing topology changes as a side effect of branch length changes (LOCAL and Continuous Change) consistently perform worse than those involving stochastic branchrearrangements (nearest neighbor interchange, subtree pruning and regrafting, tree bisection and reconnection, or subtree swapping). Among the latter, moves that use an extension mechanism to mix local with more distant rearrangements show better overall performance than those involving only local or only random rearrangements. Moves with only local rearrangements tend to mix well but have long burnin periods, while moves with random rearrangements often show the reverse pattern. Combinations of moves tend to perform better than single moves. The time to convergence can be shortened considerably by starting with a good tree but this comes at the cost of compromising convergence diagnostics based on overdispersed starting points. Our results have important implications for developers of Bayesian MCMC implementations and for the large group of users of Bayesian phylogenetics software.
L3  10.1080/10635150801886156
JF  Systematic Biology
VL  57
IS  1
SP  86
EP  103
ER 