Related Experiment Video
Updated: Jan 4, 2026
07:52
Author Spotlight: Enhancing Cryo-Electron Microscopy by Automated Data Collection and Analysis Techniques
Published on: December 1, 2023
1.4K
Virtual Screening in the Cloud: How Big Is Big Enough?
Christoph Grebner1, Erik Malmerberg1, Andrew Shewmaker2
1Hit Discovery, Discovery Sciences, BioPharmaceuticals R&D, AstraZeneca, SE-43183 Gothenburg, Sweden.
Journal of Chemical Information and Modeling
|November 5, 2019
Summary
High-throughput virtual screening using 3D similarity on cloud platforms enables searching billions of molecules. This scalable approach expands accessible chemical space for discovering novel drug candidates efficiently.
Area of Science:
- Computational Chemistry
- Drug Discovery
- Bioinformatics
Background:
- Virtual screening is crucial in Computer-Assisted Drug Design (CADD).
- Ligand-based similarity searches are common early-stage methods.
- Computational resource limitations restrict screening scale and speed.
Purpose of the Study:
- To describe a high-throughput virtual screening project utilizing 3D similarity.
- To demonstrate the scalability and flexibility of cloud computing for virtual screening.
- To explore the impact of large search spaces on hit discovery and computational costs.
Main Methods:
- Employed 3D similarity searching (FastROCS) on a cloud platform (Orion).
- Utilized automated evaluation workflows for efficient processing.
- Generated and searched billions of virtual molecules.
Main Results:
- Cloud resources enabled fully scalable and flexible virtual screening.
- Provided access to an unprecedented explicit 3D virtual chemical space.
- Investigated the relationship between search space size and novel hit identification.
Conclusions:
- Cloud-based 3D virtual screening offers a powerful, scalable solution for drug discovery.
- Expanding the search space significantly impacts hit discovery and list size.
- Computational and economic factors are key considerations for resource scaling in virtual screening.
