QSAR and scaffold-based optimization of HMGR inhibitors using cheminformatics and machine learning

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Atherosclerosis, a major cardiovascular disease risk factor, is linked to high cholesterol.
• 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR) is a key enzyme in cholesterol synthesis and a drug target.
• Current HMGR inhibitors (statins) have adverse effects, necessitating new therapeutic options.

Purpose of the Study:

• To identify and optimize novel inhibitors of HMGR using cheminformatics and machine learning.
• To develop a systematic pipeline for discovering drug-like molecules targeting HMGR.
• To improve upon existing HMGR inhibitors by enhancing safety and efficacy profiles.

Main Methods:

• Utilized a curated ChEMBL database for analysis.
• Applied physicochemical descriptor profiling, exploratory data analysis, and principal component analysis (PCA).
• Developed quantitative structure-activity relationship (QSAR) models using gradient boosting and XGBoost, achieving a cross-validated R² of 0.70.

Main Results:

• Identified active molecules clustering around complex cyclic scaffolds with aromatic and nitrogen-containing motifs.
• Optimized four promising scaffolds using ligand-based R group enumeration.
• Achieved enhanced hydrogen bonding, polarity, and multiparameter optimization (MPO) scores for identified scaffolds.

Conclusions:

• Integrated cheminformatics and machine learning provide a robust pipeline for novel HMGR inhibitor discovery.
• The study highlights promising scaffolds with optimized drug-likeness for potential therapeutic development.
• This approach offers a systematic strategy for developing safer and more effective HMGR inhibitors.