A *likelihood model of gene family evolution
The duplication of genes is a source of new genes, and thus can be of great interest. This thesis presents a simple model of gene duplication and gene loss, and a method by which the probability of a gene phylogeny can be computed for particular rates of these processes. This allows the likelihood of duplication and loss rates for a particular gene phylogeny to be determined. Using such calculations, I have examined the ability to estimate duplication and loss for a variety of different types of gene phylogenies.
Genetic data do not specify a single gene phylogeny with certainty. By integrating over many possible gene phylogenies via Markov chain Monte Carlo methods, uncertainty in the gene phylogeny is allowed. I have written a computer program to perform this calculation for DNA sequence data. With this, I have examined sources of variance in the estimation of gene duplication and loss rates.