Devising Breast Cancer Diagnosis Protocol through Machine Learning

Abstract

Breast cancer is a life threatening disease and have serious implications to health. It is further categorized on the bases of receptors including ER+ and HER2-. Breast cancer is a multifaceted disease that has many subcategories characterized by unique genetic features. This research focuses on two important subgroups of receptors, including ER+ and HER2-. We conducted an analysis of gene expression data obtained from reliable source (Array Express: E-GEOD-52194, E-GEOD-75367, and E-GEOD-58135) in order to reveal the complex molecular details of these subtypes. The computational pipeline we used identified 396 genes that exhibited distinct patterns of gene expression in ER+ and HER2- breast cancers. The diagnostic and prognostic significance of these genes was evaluated using machine learning methods, namely SVM and decision tree models. Metrics like as accuracy, sensitivity, and specificity provide insights into their usefulness. Furthermore, the use of the STRING database for network analysis revealed significant signaling pathways and biological processes associated with the development of ER+ and HER2- breast cancer. The results of our research enhance our comprehension of these subcategories, which might possibly facilitate more accurate diagnoses and focused treatment interventions. This work provides valuable information on the genetic foundations of ER+ and HER2- breast cancer, which has potential implications for enhancing patient treatment and outcomes.