Identifying Top Gene Contributors to Triple Negative Breast Cancer Health Disparities Among African American Women: A Machine Learning Approach

Abstract

Purpose: Triple Negative Breast Cancer (TNBC) is a breast cancer subtype which multiple studies have shown to be disproportionately prevalent among premenopausal African American women. The factors contributing to the TNBC health disparities remain unclear. Methods: Here, we developed a highly accurate, reproducible machine learning classification model that used patient gene expression values as predictor attributes to classify 100 TNBC patients as either African American or non-African American. Results: By using weighting methods and comparison of classification performance at varying levels of attributes, our study identified a subset of genes able to accurately classify TNBC patients by race. Intriguingly, the top genes of this subset are linked to diabetes, indicating that diabetes may associate with the TNBC health disparities. Conclusions: Our study demonstrated the factors contributing the TNBC health disparities and provided a subset of genes that may be targetable for precision medicine development to address disparity of TNBC among the African American female population.