ANALYZING LATERAL GENE TRANSFER WITH MACHINE LEARNING AND PHYLOGENETIC METHODS

Abstract

Lateral gene transfer (LGT) is an important evolutionary process. To further understand the impact of LGT, it is necessary to perform the quantitative study of LGT. In this thesis, we mainly study three related problems: how to detect large genomic regions originated from LGT (genomic islands, GIs); how to model LGT with phylogenetic networks; and how to detect LGT events. For the first problem, we developed two machine learning methods to detect GIs: GI-SVM and GI-Cluster. For the second problem, we implemented fast exponential-time programs for solving two problems on arbitrary phylogenetic networks, the tree containment problem (TCP) and the cluster containment problem (CCP). Our CCP program is further extended into a program for fast computation of the Soft Robinson--Foulds distance between phylogenetic networks. For the third problem, we conducted a case study on cyanobacterial genomes to investigate the complementary properties of different LGT detection methods.