Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.815618
Title: Bayesian methods for protein quantification in mass spectrometry proteomics
Author: Phillips, Alexander
Awarding Body: University of Liverpool
Current Institution: University of Liverpool
Date of Award: 2020
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
Current workflows in mass spectrometry proteomics are able to identify thousands of proteins in a single biological sample by breaking them down through enzymatic digestion into smaller molecules, peptides. Quantification of the differences in abundances of proteins between populations is based on measurements of these peptides at multiple charge states, features. Subsequent determination of significant changes in the abundance of proteins between those populations remains a challenge. This is complicated further by the presence of shared peptides arising from multiple proteins. This PhD thesis explores some of the statistical modelling challenges associated with the quantification of proteins in mass spectrometry proteomics. Firstly, an overview of the existing literature is presented, focusing on current methods for quantifying proteins and differential expression analysis. A current Bayesian pipeline for quantifying proteins from feature-level data using Markov chain Monte Carlo sampling is then evaluated. A method by which sampling might be made more efficient by exploiting conjugate distributions is then proposed. Similar conjugate analysis is then applied to the specific problem of differential expression analysis. The analytic nature of the resulting model is exploited to achieve fast inference without the need for computationally expensive numerical integration. This enables a model comparison approach to statistical testing, with the aim of achieving calibrated estimates of false discovery rate. A Bayesian hierarchical model is then proposed for the analysis of shared peptides. Finally, overarching conclusions are drawn and recommendations for future research are discussed.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.815618  DOI:
Share: