Date of Award
9-22-2025
Document Type
Dissertation
Degree Name
Molecular Biosciences, Ph.D.
First Advisor
Jason Causey
Committee Members
Jake Qualls; Jonathan Stubblefield; Karl Walker; Xiuzhen Huang
Abstract
The publication of the human genome in 1994 launched a computational biology revolution. Despite subsequent advances in sequencing and computing technology, the research community lacks robust and reproducible tools for interpreting differential expression of genes. Modern high-throughput genome sequencing and high-density arrays allow unprecedented computational analysis of genetic profiles. These developments combined with the rise artificial intelligence and machine learning powered by unprecedented amounts of data present new opportunities in bioinformatics. A novel gene ranking method, AUCg, is presented and applied to genetic expression data from patients with multiple myeloma. The AUCg method is compared to popular Bioconductor tools, and the corresponding advantages and drawbacks are discussed. Other recently developed tools including PyDESeq2 and Bioinfokit are also implemented, resulting in the identification of significantly over- and under-expressed genes that may serve as therapeutic targets. Finally, a review of the known roles and interactions of identified genes offers new insight to the disease biology and progression of multiple myeloma.
Rights Management
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Recommended Citation
Fowler, Jennifer M., "Machine Learning for Genetic Expression" (2025). Student Theses and Dissertations. 1106.
https://arch.astate.edu/all-etd/1106
