Comparative genomics to identify function associations

Many microbial genomes have been completely sequenced. The presence/absence data for thousands of genes across a diverse array of species allow one to quickly identify genes that are functionally associated. Because of shared ancestry among biological species, such functional association needs to be phylogenetically controlled. The computation requires the construction of a reliable species tree, map the presence/absence data onto the tree, and test two Markov chain models, one assuming no functional association (with four parameters) and the other assuming association (with eight parameters) by a likelihood ratio test. In the past, this has been done only for small data set in a semi-automated way. This project will create a computer program to automate the entire process. Students in this project will learn the general conceptual framework as well as the relevant knowledge of stochastic processes and statistical estimation methods to facilitate their future participation as graduate students.

Faculty Supervisor:

Xuhua Xia

Student:

Partner:

Discipline:

Sector:

University:

University of Ottawa

Program:

Globalink Research Internship