An Effective and Robust Approach Based on Malatya Centrality Algorithm for Interpreting Cheminformatics Graphs Using Maximum Clique

Abstract

Cheminformatics graphs are derived by transforming the atomic nodes and bonds of chemical compounds into graph structures and are used to analyze the chemical and structural properties of molecules. In this study, an effective and robust approach based on the Malatya Centrality Algorithm is proposed for identifying the maximum clique in cheminformatics graphs. The proposed method transforms cheminformatics graphs by taking their complement and calculates the Malatya centrality values for these graphs. Using these values, the minimum independent set is identified in the complemented graph, which corresponds to the set of nodes forming the maximum clique in the original graph. The study demonstrates, through tests on various cheminformatics graphs, including enzyme and molecular graphs, that maximum clique and chromatic number values provide significant insights into the structural properties of these graphs. Notably, the maximum clique value was often calculated as 2 for bipartite graphs. Additionally, it was observed that enzyme graphs exhibit maximum clique and chromatic number values that are optimal or near-optimal, with some graphs possessing perfect graph properties. The proposed approach offers an effective and robust solution for structural analysis in cheminformatics graphs.