Format

Send to

Choose Destination
Syst Biol. 2015 Sep;64(5):741-51. doi: 10.1093/sysbio/syv030. Epub 2015 May 25.

Monte Carlo Strategies for Selecting Parameter Values in Simulation Experiments.

Author information

1
Department of Mathematics and Statistics, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand jleigh@maths.otago.ac.nz.
2
Department of Mathematics and Statistics, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand.

Abstract

Simulation experiments are used widely throughout evolutionary biology and bioinformatics to compare models, promote methods, and test hypotheses. The biggest practical constraint on simulation experiments is the computational demand, particularly as the number of parameters increases. Given the extraordinary success of Monte Carlo methods for conducting inference in phylogenetics, and indeed throughout the sciences, we investigate ways in which Monte Carlo framework can be used to carry out simulation experiments more efficiently. The key idea is to sample parameter values for the experiments, rather than iterate through them exhaustively. Exhaustive analyses become completely infeasible when the number of parameters gets too large, whereas sampled approaches can fare better in higher dimensions. We illustrate the framework with applications to phylogenetics and genetic archaeology.

KEYWORDS:

Importance sampling; Markov chain Monte Carlo; phylogenetic analysis; plant domestication; simulation

PMID:
26012871
DOI:
10.1093/sysbio/syv030
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Silverchair Information Systems
Loading ...
Support Center