Structure-based quantitative structure--activity relationship modeling of estrogen receptor β-ligands

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Estrogen receptor (ER) β-selective ligands are explored for treating neurodegenerative diseases.
• Structure-activity relationship (SAR) and X-ray data for ERβ ligands are available.
• Developing robust quantitative structure-activity relationship (QSAR) models is vital for advancing ERβ ligand development.

Purpose of the Study:

• To build predictive QSAR models for ERβ-selective ligands.
• To leverage both SAR data and 3D structural information of ERβ.
• To facilitate the discovery of novel ERβ-selective molecules.

Main Methods:

• Employed a novel structure-based QSAR method (structure-based pharmacophore keys QSAR).
• Utilized existing SAR data and 3D structural information of the ERβ ligand-binding domain.
• Applied a robust QSAR workflow to analyze 37 ERβ ligands.

Main Results:

• Generated four sets of QSAR models.
• Approximately 30 models demonstrated high predictive performance (training-r² > 0.60 and test set-R² > 0.60).
• Validated the efficacy of the structure-based QSAR approach.

Conclusions:

• Developed an ensemble of predictive models for ERβ ligands.
• These models will aid in the future discovery of novel ERβ-selective compounds.
• The study highlights the utility of structure-based QSAR in drug design.