UC Santa Barbara

Proteins represent a major component of nearly all biological processes, including metabolic reactions, cellular structure, and immune function. It is of interest to understand the utility of mRNA as proxy for protein abundance, as mRNA transcription represents a critical stage in the process of protein synthesis, and can be quantified with fewer technical challenges relative to measuring the proteome; however, its evaluation requires quantification of the role that post transcriptional regulation plays in protein synthesis, and the degree to which it influences relative protein abundance within individual genes. In this dissertation we explore the effect of post transcriptional regulation using high throughput mRNA and protein measurements, collected for individual tissues and cells, by estimating the relative protein-to-mRNA ratio (rPTR). Analysis of high throughput mRNA and protein measurements requires careful modeling of noise and an understanding of technical factors that contribute to measurement bias. We estimate rPTR using a Bayesian hierarchical model that directly models technical factors, relevantbiological processes, and sampling variability. This model presents a useful approach to estimate rPTR at the gene level and identify genes and functional sets of genes that have extreme rPTR in measured tissues and cell types. Finally we evaluate the effect of genelevel technical factors, including batch effect and sequencing depth, and their influence of rPTR estimation by applying our Bayesian model to adjusted transcript counts using two technical-effect correction methods