Machine learning-based analysis of electronic properties as predictors of anticholinesterase activity in chalcone derivatives

In this study, we investigated the correlation between the electronic properties of anticholinesterase compounds and their biological activity. While this correlation has been effectively explored in previous studies, we employed a more advanced approach: machine learning. We analyzed a set of 22 molecules sharing a similar chalcone skeleton, categorizing them into two groups based on their IC₅₀ indices: high and low activity. Using the open-source software Orca, we calculated the geometries and electronic structures of these molecules. Over a hundred parameters were extracted, including Mulliken and Lowdin electronic populations, molecular orbital energies, and Mayer’s free valences, forming the foundation for machine learning features. Through our analysis, we developed models capable of distinguishing between the two groups. Notably, the most informative descriptor relied solely on electronic populations and orbital energies. Identifying computationally relevant properties for biological activity enhances drug development efficiency, saving time and resources.