Date Approved

5-26-2016

Embargo Period

11-28-2016

Document Type

Thesis

Degree Name

M.S. Bioinformatics

Department

Biological Sciences

College

College of Science & Mathematics

Advisor

Hickman, Mark

Committee Member 1

Breitzman, Anthony

Committee Member 2

Caputo, Gregory

Keywords

Artificial Neural Network, Bootstrapping, Evolutionary constraints, Gene Expression, RNA-seq, Saccharomyces cerevisiae

Subject(s)

Bioinformatics; Gene expression

Disciplines

Bioinformatics

Abstract

It has been widely reported that some genes are expressed at a higher level than others. However, it has not been shown whether each gene is expressed consistently between studies and at the same level compared to all other genes. Here, we examined six RNA-seq datasets and found that the mRNA level of each gene is indeed consistent relative to all other genes. This result implies that there are evolutionary pressures that drive genes to maintain either low or high expression. In order to identify these pressures, we compared gene expression level to the features of each gene (or associated protein), such as biological function, molecular process, or localization. We found many possible pressures; for example, genes involved in translation and the ribosomal processes were expressed at high levels while genes involved in transcription and DNA-related processes were expressed at low levels. Furthermore, through the optimization of an artificial neural network, we were able to use several of these features to predict gene expression with 65-75% accuracy. In conclusion, these results show that gene expression level is controlled by several evolutionary constraints, including biological function, molecular process, and cellular localization.

Download

Share

COinS