Related Experiment Video
Updated: Jul 6, 2026
08:35
Achieving Efficient Fragment Screening at XChem Facility at Diamond Light Source
Published on: May 29, 2021
SMASH: Screening Molecules Accurately on Small Hardware. Fast, user-friendly, enhanced with a machine learning
Alfredo Suárez-Alonso1, Leonardo D Herrera-Zúñiga2,3, Mayra Lozano-Espinosa4
1Laboratorio de Farmacología, Departamento de Ciencias de La Salud, Universidad Autónoma Metropolitana-Iztapalapa, Ciudad de México, México.
Journal of Molecular Modeling
|July 4, 2026
Summary
SMASH is a new, free tool that accelerates molecular docking and virtual screening using machine learning and GPU acceleration. It accurately predicts binding sites and ligand poses, aiding drug discovery research.
Area of Science:
- Computational chemistry
- Drug discovery
- Bioinformatics
Background:
- UAM-Ixachi was previously released to simplify molecular docking.
- SMASH builds upon UAM-Ixachi with significant upgrades for enhanced performance.
Purpose of the Study:
- To introduce SMASH, an upgraded, user-friendly tool for molecular docking and virtual screening.
- To leverage machine learning and GPU acceleration for faster and more accurate predictions.
Main Methods:
- Implemented PDB2PQR, P2Rank, MGL Tools, OpenBabel, AutoDock-GPU, Vina-GPU, and SCORCH.
- Utilized MMFF94, AD4, and Vina force fields for simulations.
- Integrated machine learning for binding site prediction and data clustering using K-means.
Main Results:
- SMASH accurately reproduces ligand poses from Protein Data Bank complexes in automatic mode.
- The tool successfully differentiates active and decoy ligands from a DUD-E dataset.
- SMASH handles large projects efficiently using local computational resources.
Conclusions:
- SMASH offers a powerful, efficient, and accurate solution for molecular docking and virtual screening.
- The tool democratizes advanced computational chemistry methods for broader research accessibility.
