Optimization, random resampling, and modeling in bioinformatics

Quantitative phenotypes regulated by multiple genes are prevalent in nature and many diseases falls into this category. High-throughput sequencing and high-performance computing provides a basis to understand quantitative phenotypes. However, finding a statistical approach correctly model the phenotypes remain a challenging problem. In this work, I present a resampling-based approach to obtain biological functional categories from gene set and apply the approach to analyze lithium-sensitivity of neurological diseases and cancer. Then, the non-parametrical permutation-based approach is applied to evaluate the performance of a GWAS modeling procedure. While the procedure performs well in statistics, search space reduction is required to address the computation challenge.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Biophysics & Computnl Biology

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2018

Author dc:creator

• Ge, Weihao

Contributors dc:contributor

• Jakobsson, Eric
• Mainzer, Liudmila S
• Sinha, Saurabh
• Nelson, Mark
• McHenry, Kenton

• resampling
• gene ontology
• pathway
• FDR
• GWAS
• epistasis
• search space reduction

Statement dc:rights

• Copyright 2018 Weihao Ge

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/101707

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/101707