COMBINE: A Novel Ensemble Learning Model for Improved Breast Cancer Prognosis in African American Women through Multi-Omics Data Integration

Abstract

AbstractBreast cancer exhibits a disproportionate impact on African American women below 50 years of age, as they encounter elevated incidence rates, more aggressive cancer subtypes, and increased mortality in comparison to other racial and ethnic groups. To enhance the prediction of onset risk and facilitate timely intervention and treatment, it is imperative to examine the underlying genetic and molecular factors associated with these disparities. In this study, we introduce an innovative ensemble learning model, termed COMBINE, which amalgamates three disparate types of omics data to augment the precision of breast cancer prognosis classification and diminish the model's time complexity. A comparative analysis of the fusion effects for African American and White women reveals a substantial improvement in the fusion effect for African American women. Moreover, gene enrichment analysis underscores the significance of race in selecting pertinent biomarkers. To address multiobjective problems in cancer prognosis classification, we employ a combination of qualitative and quantitative methodologies, along with ensemble learning. This multifaceted approach enables the exploration of novel concepts for multi-omics data applications, potentially leading to more customized and efficacious treatment strategies. The insights derived from this study emphasize the potential of ensemble learning as a multi-omics data fusion technique, specifically in the context of its application in cancer prognosis classification. By refining our comprehension of the genetic and molecular factors contributing to the disparities in breast cancer incidence and outcomes, we can ultimately improve healthcare outcomes for African American women and alleviate the burden of this formidable disease.