Graduation Year

2024

Date of Submission

4-2024

Document Type

Open Access Senior Thesis

Degree Name

Bachelor of Arts

Department

Biology

Reader 1

Erin Jones

Reader 2

Kyle Jay

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2024 Javier Castillo

Abstract

The following proposal describes a modular machine learning approach that detects malfunctioning genes and pathways in cancer using the transcriptome of cancer patients. The transcriptome is underused in precision oncology and, combined with machine learning, can aid in the identification of hidden responders. Applied to the PI3K/AKT/mTOR pathway, this method can be used to predict PIK3CA functional status and identify phenocopying variants of deleterious PIK3CA mutations. The classifier will do this by integrating RNA-seq, copy number, and mutation data from tumors to determine the functional status of PIK3CA using a set of learned gene-specific weights. The classifier will then be applied to cell line datasets with pharmacological profiling data on Buparlisib and Copanlisib to determine if there is a correlation between classifier scores and sensitivity to PI3K inhibitors. If successful, these classifiers have potential for identifying phenocopying events, suggesting their usefulness as a biomarker application to potentially reveal hidden responders that may have otherwise been missed by sequencing. Ultimately, as datasets expand and algorithms improve, our capacity to pinpoint comprehensive treatment strategies that address the specific weaknesses of each tumor will enhance significantly.

Share

COinS