QSAR and Scaffold-Based Optimization of HMGR Inhibitors via Cheminformatics and ML

A curated dataset from the ChEMBL database used to identify and optimize potential 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR) inhibitors. The data was analyzed through physicochemical descriptor profiling, exploratory data analysis, and principal component analysis, leading to the development of quantitative structure–activity relationship (QSAR) models. The dataset was authored by Priya Antony and last updated on 2026-04-30.

Use Cases

• Developing QSAR models for HMGR inhibition based on physicochemical descriptors mentioned in the description
• Optimizing drug-like molecular scaffolds based on Murcko scaffold extraction and R group enumeration described in the study
• Training gradient boosting and XGBoost regressors for activity prediction as demonstrated in the research
• Analyzing molecular clustering patterns around complex cyclic frameworks enriched in aromatic and nitrogen-containing motifs

Strengths

• Dataset is curated from the authoritative ChEMBL database
• Study reports a tuned XGBoost model achieving a cross-validated R² of 0.70
• Four scaffolds were successfully optimized with improved multiparameter optimization (MPO) scores

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• The primary data file is a 1.5 MB PDF, which may not be in a directly machine-readable tabular format

Provenance

Source

ChEMBL database

Collection Method

Curated and analyzed through cheminformatics and machine learning approaches

Freshness

Last updated 2026-04-30 05:38:41; freshness should be verified