PSB 2002 Tutorial

 PSB 2002

Molecular Evolution and Phylogenetic Analysis

David Pollock and Richard Goldstein

Description

Increasingly, biocomputing is taking advantage of the methods and approaches of evolutionary biology. This is due to the intersection of a number of different developments: the availability of whole genomes from a growing number of organisms, the availability of high-speed computational facilities that allow sophisticated computational and statistical models, and the growing realization of the power of comparative sequence analysis. Such approaches require an understanding of the ways that the corresponding organisms evolved and changed. There are, unfortunately, numerous misconceptions about the way that evolution occurs, misunderstandings about current evolutionary theory, and a lack of awareness of available methods and software.

This tutorial will focus on giving participants a broad overview of the nature of molecular evolution, techniques that have been developed for modeling this evolution, tools to interpret biological information in an evolutionary context, and software that has and is being developed to analyze evolutionary processes.

We will discuss the following topics:

I. The underlying process of evolutionary change

The manner in which evolution proceeds. Mutation, fixation, and substitution. The difference between genetic change and genomic change. Various models of the substitution process at both the nucleic acid and protein level, as developed by Jukes and Cantor, Kimura, and others.

II. Recent theories of evolution

Eigen and the idea of pseudospecies; the neutral theory and near-neutral theory, neutral networks and their role in evolutionary dynamics. Selection vs. neutrality. Fitnesses and fixation rates.

III. Approaches to phylogenetic analysis

Issues involving likelihood (maximum likelihood and Bayesian inference), parsimony, and distance methods -- when and why are they justified, under what circumstances are they appropriate, when do they have problems (such as long-branch attraction, recombination). Typical model assumptions and extensions towards biological reality. Molecular clocks and modifications. Amino acid models, codon models and synonymous and non-synonymous substitution rates. Adaptation, coevolution, and ancestral reconstruction.

IV. Algorithms in phylogenetic analysis

Bootstrapping and jack-knifing. Neighbor-joining, UPGMA, quartet-puzzling, and mutation data matrices. Tree space searching algorithms. Markov-Chain Monte Carlo (MCMC) and importance sampling.

V. Available Software

PAUP*, Phylip, PAML, MOLPHY, MrBayes.

Biographical Sketches:

David Pollock is Assistant Professor in the Department of Biological Sciences and the Biological Computation and Visualization Center at Louisiana State University. His research focus is on the relationship between protein structure and function and sequence variation, particularly the effect of structural context on coevolution between residues.

Richard Goldstein is Associative Professor in the Department of Chemistry and Biophysics Research Division at the University of Michigan. His research focuses on how we can understand the properties of biomolecules in their evolutionary context, how we can use the evolutionary record to understand specific proteins, and how we can use the properties of proteins to understand the evolutionary record.

Back to the main PSB page Updated: August 1, 2001