Novel computational methods for discordance-aware phylogenomic analysis

Tabatabaee, Seyedeh Yasamin

Permalink

https://hdl.handle.net/2142/132481 Copy

Description

Title

Novel computational methods for discordance-aware phylogenomic analysis

Author(s)

Tabatabaee, Seyedeh Yasamin

Issue Date

2025-11-11

Director of Research (if dissertation) or Advisor (if thesis)

Warnow, Tandy

Doctoral Committee Chair(s)

Warnow, Tandy

Committee Member(s)

• El-Kebir, Mohammed
• Gropp, William
• Liu, Ge
• Mirarab, Siavash

Department of Study

Computer Science

Discipline

Computer Science

Degree Granting Institution

University of Illinois Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• phylogenetics
• phylogenomics
• gene tree discordance
• species tree estimation
• multi-species coalescent

Abstract

Inferring the evolutionary history of a set of species is a key step in many biological and medical research projects, as species trees provide a context in which problems in comparative genomics, biodiversity, phylogeography and epidemiology can be addressed. Recent advances in sequencing technologies have led to an increasing availability of genome-scale data, and today phylogenomics projects construct species trees using hundreds to thousands of loci, potentially whole genomes. However, species tree estimation from multi-locus datasets presents several statistical and computational challenges, as most problems in this area are NP-hard. Also, due to a phenomenon known as “gene tree heterogeneity”, different locations within the genome of a species can evolve differently due to biological processes such as incomplete lineage sorting, gene duplication and loss and horizontal gene transfer, that further complicate species tree estimation. Despite advances in developing methods that can estimate an unrooted and non-parameterized topology of a species tree in the presence of gene tree discordance, less attention has been paid to estimating the root location, quantifying branch lengths in units that are usable for downstream analysis, and estimating divergence times. All of these are necessary for many applications of phylogenomics, such as constructing the tree of life and analyzing the origins of diseases, such as HIV and COVID-19. In this dissertation, we introduce new computational methods developed for these tasks, collectively referred to as "post-species tree analysis", that address different sources of gene tree discordance. For these methods, we present rigorous theoretical results including proofs of statistical consistency, sample complexity, and running time analyses, as well as extensive empirical results on simulated and biological datasets ranging from the root of the tree of life to recent speciations. Overall, these methods provide high accuracy and scalability for estimating the root, branch lengths and divergence times in the presence of gene tree discordance, and some are accompanied with strong theoretical guarantees.

Graduation Semester

2025-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/132481

Copyright and License Information

Copyright 2025 Seyedeh Yasamin Tabatabaee

Owning Collections