A Novel Hybrid Approach for Biological Network Alignment: Integrating Functional and Topological Information

Biological network alignment represents a fundamental computational challenge in systems biology, enabling the transfer of functional knowledge across species through the identification of conserved network regions. Traditional alignment methods face significant limitations when dealing with dynamic networks and large-scale biological data. This paper presents Wave+, a novel hybrid alignment strategy that combines topological and functional approaches to overcome these challenges. Our method leverages protein interaction data from the STRING database to enhance alignment quality through the integration of both structural and functional information. Wave+ demonstrates superior performance across multiple evaluation metrics, including node conservation (P-NC, R-NC, F-NC), edge conservation (GS3), and overall alignment quality (NCV, NCV-GS3). Experimental validation on Yeast protein-protein interaction networks (Yeast0, Yeast5, and Yeast25) shows consistent improvements over existing methods, with F-NC scores reaching 90% on noisy datasets. The hybrid nature of our approach makes it particularly suitable for large-scale biological networks while maintaining computational efficiency.