Links between maximum likelihood and maximum parsimony under a simple model of site substitution

  title={Links between maximum likelihood and maximum parsimony under a simple model of site substitution},
  author={Christopher P. Tuffley and Mike A. Steel},
  journal={Bulletin of Mathematical Biology},
Stochastic models of nucleotide substitution are playing an increasingly important role in phylogenetic reconstruction through such methods as maximum likelihood. Here, we examine the behaviour of a simple substitution model, and establish some links between the methods of maximum parsimony and maximum likelihood under this model. 
Sufficient Conditions for Two Tree Reconstruction Techniques to Succeed on Sufficiently Long Sequences
  • M. Steel
  • Biology, Mathematics
    SIAM J. Discret. Math.
  • 2001
Simple sufficient conditions for two tree reconstruction methods (maximum parsimony and maximum compatibility) to correctly reconstruct a tree when applied to sufficiently many sequence sites generated under a simple stochastic model are described.
A Bayesian perspective on a non-parsimonious parsimony model.
A Bayesian approach is taken to the no-common-mechanism model and place independent gamma prior probability distributions on the branch-length parameters, which allowed us to implement an efficient Markov chain Monte Carlo method for exploring the space of phylogenetic trees.
Probabilistic methods outperform parsimony in the phylogenetic analysis of data simulated without a probabilistic model
The Bayesian Mk model emerges as the most reliable approach for categorical morphological analyses and consistent datasets render method choice largely irrelevant, as all methods perform well with high consistency datasets, but the largest discrepancies in accuracy occur with low consistency datasets (high homoplasy).
A likelihood approach to estimating phylogeny from discrete morphological character data.
Several new avenues of research are opened by an explicitly model-based approach to phylogenetic analysis of discrete morphological data, including combined-data likelihood analyses (morphology + sequence data), likelihood ratio tests, and Bayesian analyses.
A Likelihood Approach to Estimating Phylogeny from Discrete Morphological Character Data
This paper explores the possibility of using standard, well-behaved Markov models for estimating morphological phylogenies (including branch lengths) under the likelihood criterion.
Most Parsimonious Likelihood Exhibits Multiple Optima for Compatible Characters
It is shown that, despite exhibiting behavior similar to parsimony, multiple local optima can occur for compatible characters for the most parsimonious likelihood estimator, and thorough understanding of likelihood criteria is necessary before they are used to analyze biological data.
Comparative evaluation of maximum parsimony and Bayesian phylogenetic reconstruction using empirical morphological data
The results, which were based on a large set of empirical matrices, corroborate recent findings that BI is less precise than MP and show that differences between both approaches were not influenced by increasing sample size.
Quantifying the accuracy of ancestral state prediction in a phylogenetic tree under maximum parsimony
The reconstruction accuracy of this estimation method is investigated further, under a simple symmetric model of state change, and a number of new results are obtained, both for 2- state characters, and r-state characters ($$r>2$$ r>2).
Surprising properties of Maximum Parsimony on phylogenetic networks
It is shown that the various definitions Maximum Parsimony on networks can also be proven to be equivalent to certain functions which are similar to the likelihood concept, and their biological meaningfulness is discussed.