DISSECTING PATHWAYS WITH THE YEAST KNOCKOUT COLLECTION

Abstract

The yeast knockout collections provide opportunities to perform massively parallel phenotyping of deletion mutants for almost every yeast open reading frame. I used the knockout collection to screen for synthetic lethal partners, defined as alleles that cause lethality when combined but are nonlethal alone, with CTF4 and CTF18 and present the results in Chapter 2. I developed procedures for interpreting microarrays designed to compare changes in oligonucleotide TAGs specific to each knockout strain and present those methods in Chapters 3 and 4. These TAG microarrays allow thousands of experiments to screen for synthetic lethality among pairs of null alleles to be accomplished relatively quickly. In Chapter 4, I present 1410 novel predicted synthetic lethal interactions based on 707 currently completed screens. Interpretation of synthetic lethality is presented with a computational approach in Chapter 5, termed the congruence score. High congruence scores associate genes into common pathways, and I use the method to predict that YLL049W is a component of the dynein-dynactin nuclear orientation pathway. In Chapter 6, I propose a generalization of the congruence score to any phenotype, such as growth rate in the presence of various compounds, or even nonquantitative phenotypes such as cell morphology. This procedure connects genes based on similarity of multiple phenotypes using an application of information theory to produce a shared information score. Using gene ontology similarity, I show that high scores are associated with similarly annotated genes.