Integrative graph theory and boolean modeling for breast cancer network reconstruction in precision medicine

Abstract

Breast cancer comprises complex molecular interactions that can be represented as biological networks. Understanding these networks is essential for identifying regulatory hubs and potential therapeutic targets in precision oncology. This study reconstructed a breast cancer protein–protein interaction (PPI) network using data from STRING-DB, KEGG, and SIGNOR. Graph theory was applied to compute topological metrics—degree, betweenness, and clustering coefficients—to identify key proteins and functional modules, while the Markov Cluster Algorithm (MCL) detected community structures. Boolean modeling simulated network dynamics by binarizing interaction strengths at a confidence threshold of 0.7. The reconstructed network contained 150 nodes and 1,359 edges, exhibiting a scale-free topology (γ = 2.1) and modular organization (global clustering coefficient 0.522). BRCA1 and TP53 emerged as densely connected hubs, whereas EGFR and AKT1 acted as major signaling conduits linking multiple pathways. MCL revealed four primary clusters associated with DNA repair, cell-cycle regulation, growth signaling, and survival pathways. Boolean simulations demonstrated that perturbing these hub proteins significantly altered network states linked to proliferation and apoptosis resistance. Notably, TP53 restoration was predicted to stabilize Basal-like breast cancer networks, while inhibition of AKT1 or EGFR suppressed pro-proliferative attractors. Integrating graph theory with Boolean modeling thus provides a systems-level framework for understanding molecular regulation in breast cancer. The identification of BRCA1, TP53, EGFR, and AKT1 as high-centrality nodes highlights their importance as potential therapeutic targets and supports the advancement of precision medicine approaches tailored to breast cancer network dynamics.