Bayesian inference for the analysis of RNA-Seq data

Hoerbst, Franziska (2024) Bayesian inference for the analysis of RNA-Seq data. Doctoral thesis, University of East Anglia.

Preview

PDF Download (22MB) | Preview

Abstract

This thesis explores the application of Bayesian inference in the analysis of RNA-Sequencing data. We discuss the foundational principles of science and the mathematical articulation of belief updating via probability theory. In the Introduction we become familiar with the elegance and effectiveness of Bayesian approaches in addressing both simple examples and experimental research problems (Chapter 1). After introducing the RNA-Sequencing (RNA-Seq) technique and data (Chapter 2), we examine two use cases of Bayesian inference in the analysis of such data. First, we explore Bayes factors as a means of quantifying the evidence for gene expression changes (Chapter 3). Second, we propose Bayes factors to evaluate the evidence in RNA-Seq data for long-distance mobile messenger RNAs in plants (Chapter 6). For both applications, we perform in-depth analyses of the data (Chapters 4 and 5), test the methods and assumptions on simulated data, and compare it to currently popular published solutions (Chapters 3 and 6). We could confirm the outstanding performance of Bayes factors for the analysis of simulated and real data. Furthermore, we present and propose further inquiries, that uncover limitations and obstacles given by data collection and processing which cannot be addressed in statistical analyses (Chapters 7 and 8). The work in this thesis underscores the huge potential of Bayesian inference in enhancing scientific understanding and addresses the complexities involved in RNA-Seq data interpretation. The presented findings provide concise solutions for two statistical problems in RNA-Sequencing data analyses advancing our abilities to communicate new learnings from data in this field.

Item Type:	Thesis (Doctoral)
Faculty \ School:	Faculty of Science > School of Biological Sciences
Depositing User:	Jennifer Whitaker
Date Deposited:	25 Apr 2025 15:45
Last Modified:	25 Apr 2025 15:45
URI:	https://ueaeprints.uea.ac.uk/id/eprint/99094
DOI:

Downloads

Actions (login required)

View Item