Parallel ACO-VNS for the solution of binary combinatorial optimization problems in computational systems biology

In this work, a parallel hybrid framework that combines Ant Colony Optimization (ACO) with Variable Neighborhood Search (VNS) to enhance both exploration and intensification is proposed. The method integrates a multicolony architecture with a self-adaptive cooperation mechanism that dynamically regulates information exchange among colonies during runtime. This design aims to overcome the limitations of standalone ACO while maintaining scalability on high-performance computing platforms. The proposed approach is evaluated on a set of challenging large-scale binary optimization benchmarks derived from cell signaling network modeling. Experimental results show that the ACO–VNS hybrid consistently outperforms standalone ACO. The results demonstrate the effectiveness and general applicability of the proposed framework for complex binary combinatorial optimization problems. • A new ACO proposal for binary optimization problems in the field of Biology Systems. • Hybrid ACO metaheuristic with Variable Neighborhood Search as local search. • Self-adaptive cooperation on a multicolony parallel strategy. • Tested using problems related to modelling of cell signalling networks.