Computational Prioritization of Multi-Target Inhibitors: Explainable QSAR and Docking- Based Discovery of Dual AChE/BACE1 Chemotypes

The discovery of dual acetylcholinesterase (AChE) and β-secretase (BACE1) inhibitors remains a promising strategy against multifactorial Alzheimer’s disease. Here, rigorously curated ChEMBL-derived data were used to develop explainable QSAR models for dual-inhibition prioritization. Molecules were standardized, near-duplicates were removed using a Tanimoto similarity threshold (≥ 0.80), and physicochemical outliers were filtered prior to modeling. Multiple classifiers (including LightGBM, XGBoost, Random Forest, SVM, kNN, MLP, and AdaBoost) and fingerprints (e.g., RDKit-FP, ECFP6) were benchmarked under scaffold-based nested cross-validation to prevent data leakage. Class imbalance was handled with SMOTETomek applied strictly within training folds. Model selection relied on a weighted composite score combining F1, PR-AUC, MCC, and Recall, and performance was accompanied by bootstrap confidence intervals, calibration curves, and Y-randomization controls. The top model (LightGBM + RDKit-FP) achieved strong generalization (Accuracy ≈ 0.92, Recall ≈ 1.00, PR-AUC ≈ 0.86, ROC-AUC ≈ 0.96). SHAP analysis highlighted aromatic and hydrogen-bonding substructures as key positive contributors. Prospective candidates (e.g., CHEMBL5082250, CHEMBL1651126, CHEMBL1651127) were evaluated by active-site-focused docking against AChE (PDB: 4EY7) and BACE1 (PDB: 2G94) with essential waters retained; docking scores (ΔG, kcal·mol⁻¹) were used for relative ranking only and were further consensus-rescored. SwissADME/pkCSM profiling suggested CNS-relevant properties (e.g., MPO, logBB, P-gp liability) and acceptable oral drug-likeness. Collectively, the workflow provides a reproducible and transparent pipeline for prioritizing dual AChE/BACE1 chemotypes and nominates testable scaffolds for experimental validation.