Leveraging Predictive Analytics To Improve Prognostic Models For Uterine Cancer

Uterine Corpus Endometrial Carcinoma (UCEC) presents notable disparities in incidence and outcomes across racial groups, warranting deeper investigation through transcriptomic and predictive modeling approaches. This thesis presents a comprehensive transcriptomic analysis of RNA-Seq data from 177 individuals, comprising both tumor and control samples, specifically from the UCEC_CN_High molecular subtype. The cohort was stratified by race, focusing on differences between Black and White individuals to explore race-associated gene expression patterns. To uncover genes with prognostic significance, LASSO (Least Absolute Shrinkage and Selection Operator) regression was applied for feature selection, identifying a subset of genes most strongly associated with overall survival. These selected genes were then utilized in a Cox Proportional Hazards Model to estimate patient-specific risk scores. The dataset was divided into a training set (70%) for model development and a testing set (30%) for performance evaluation. The integration of transcriptomic profiling with survival analysis enables a biologically informed risk stratification, providing insight into molecular drivers of UCEC and potential race-specific biomarkers. The findings highlight the potential of using survival modeling in cancer genomics to enhance prognostic accuracy and address health disparities. This study contributes to the growing body of research focused on personalized cancer therapy and racial equity in biomedical outcomes.