The First Report on Machine Learning-Driven q-RASAR Modeling for Predicting α7nACh Receptor Agonists: A Computational Approach for Anti-Alzheimer's Drug Discovery

In the present research, we have explored an advanced computational technique known as the quantitative Read-Across Structure-Activity Relationship (q-RASAR) framework, leveraging machine learning (ML) to enhance predictive precision. Our primary objective was to develop a statistically robust ML-based q-RASAR model for predicting the agonistic activity of compounds targeting the α7 nicotinic acetylcholine (α7nACh) receptor, a critical target in Alzheimer's disease (AD) therapy due to its role in cognitive function and neuroprotection. To achieve this, we have developed a well-validated univariate ML-based q-RASAR MLR regression model using a large dataset of 1,727 structurally diverse heterocyclic and aromatic hydrocarbon compounds, sourced from the freely accessible Binding Database (www.bindingdb.org). The developed q-RASAR model was further applied to assess the applicability domain of the Mcule database (accessible at https://mcule.com/database/), which comprises 1,91,94,405 chemical compounds, facilitating the identification of structurally relevant candidates with potential α7nACh receptor agonistic activity. A graded set of candidate molecules was proposed for further experimental validation to evaluate their potential as anti-Alzheimer’s agents. To further enhance predictability, we have developed various other ML-based q-RASAR models. Furthermore, molecular docking analysis and Molecular dynamic simulation for 100ns were conducted to investigate interactions between the target protein and ligand. The insights gained from this study underscore the crucial roles of hydrophobicity, electronic effects, degree of ionization, and steric factors as key determinants for potential anti-Alzheimer’s agents. This research not only advances our understanding of the molecular determinants influencing CNS drug permeability but also provides a valuable framework for designing next-generation anti-Alzheimer's agent. This approach integrates computational efficiency with predictive accuracy, offering a valuable framework for accelerating the discovery of novel therapeutic leads for AD.