Application of evolutionary algorithms to optimization of Boolean models in biomedicine

Boolean networks have become essential tools for modeling gene regulatory systems and understanding cellular decision-making processes, but their optimization for biological relevance remains challenging. This study presents a comprehensive benchmark comparison of three prominent Boolean network optimization methods--Caspo, MEIGO, and CellNOpt (ILP and GA)--evaluating their performance across structural robustness, method reliability, and biological relevance using Mean Squared Error (MSE) as the primary optimization criterion. Through systematic analysis of network reconstruction under varying perturbation levels (10-90%), we demonstrate that each method exhibits distinct performance profiles: CASPO achieves optimal topological similarity with computational efficiency, ILP produces reasonable MSE minimization but with high variance, while GA shows superior functional reconstruction stability despite longer computational requirements. Our results reveal critical limitations in current evaluation approaches, particularly the insufficient discriminatory power of F1 scores and Hamming distance metrics, and highlight fundamental trade-offs between data fitting accuracy and topological preservation in extended simulations. The analysis demonstrates that no single optimization method dominates across all criteria, with all methods showing significant performance degradation at perturbation thresholds above 10-30%, suggesting that method selection should be application-specific and guided by requirements for computational efficiency, reconstruction accuracy, and robustness to structural uncertainty in prior knowledge networks.