Intersecting Genomes: A Cross-Trait Meta-Analysis of Shared Heritable Risk in Triple Negative Breast Cancer and Prostate Cancer

Abstract Background: Triple-Negative Breast Cancer (TNBC) and Prostate Cancer (PCa) are distinct malignancies with emerging evidence of shared genetic predisposition. Epidemiological studies suggest familial co-aggregation and overlapping risk pathways, particularly involving androgen receptor (AR) signaling and DNA repair mechanisms. Objective: To identify shared genetic variants and biological pathways between TNBC and PCa through cross-trait genome-wide meta-analysis and functional validation. Methods: We conducted a cross-trait meta-analysis using genome-wide association study (GWAS) summary statistics for TNBC and PCa, derived from the Breast Cancer Association Consortium, PRACTICAL consortium, and UK Biobank. Genetic correlation was assessed using Linkage Disequilibrium Score Regression (LDSC). Shared loci were identified using PLACO and MTAG methods. Functional annotation and pathway enrichment were performed via FUMA, DEPICT, GSEA, and IPA. Colocalization (COLOC) and Mendelian Randomization (MR) analyses assessed causality. Machine learning classifiers (Random Forest, SVM) trained on transcriptomic data validated biological relevance. Results: A significant positive genetic correlation (rg = 0.27, p = 1.2 × 10⁻⁴) was found. Twelve genome-wide significant loci were shared between TNBC and PCa, including CYP24A1, AR, PCSK9, and BRCA1. Enriched pathways included AR signaling, DNA repair, and Notch signaling. Colocalization supported shared causal variants at 7 loci. MR indicated a unidirectional causal effect from PCa liability to TNBC risk. Classifiers achieved AUCs of 0.89 (RF) and 0.86 (SVM). Conclusion: Our findings reveal a moderate but meaningful shared genetic architecture between TNBC and PCa, supporting cross-cancer genetic screening and integrative risk assessment strategies. Keywords: TNBC, Prostate Cancer, Genetic Correlation, Cross-trait Meta-analysis, Machine Learning