Simulation · Scripps Research
AutoDock-GPU
GPU-accelerated molecular docking tool for high-throughput virtual screening of compound libraries.
What AutoDock-GPU is
AutoDock-GPU is a molecular docking application developed by Scripps Research for virtual screening campaigns. It adapts traditional molecular docking searches to execute across graphics processing units (GPUs), enabling rapid evaluation of large small-molecule collections against target protein structures.
The software is designed for screening libraries containing up to millions of compounds, prioritising computational throughput over per-pose structural accuracy. The resulting top-ranked candidate molecules can be passed downstream to tools such as DiffDock and Boltz-2 for structural refinement.
How it works
AutoDock-GPU implements a massively parallel Lamarckian genetic algorithm to sample ligand poses and evaluate interactions with a target receptor. The search procedure is parallelised across GPU hardware to accelerate docking calculations across large numbers of small molecules.
Key facts
- Version
- 1.5.3
- Domain
- Simulation
- Task
- Virtual screening
- Algorithm
- Lamarckian genetic algorithm
- GPU memory
- 40 GB
What you can use it for
- High-throughput virtual screening
- Evaluating large chemical libraries of up to a million compounds against a receptor structure to identify potential binding candidates.
- Lead filtering for downstream refinement
- Generating a ranked subset of high-scoring compounds to feed into structural refinement tools such as DiffDock and Boltz-2.
Run it here
No installation and no GPU of your own: fill in the inputs and it runs on our servers. Results stay in your browser unless you sign in.
Not runnable yet
This model isn't available to run right now.Run this model
AutoDock-GPU
Strengths and limits
Good at
- Accelerates classical docking search procedures through massively parallel execution on GPU hardware.
- Enables high-throughput evaluation of large libraries containing up to one million compounds.
Watch out for
- Prioritises screening throughput over per-pose docking accuracy.
- Requires substantial hardware resources, needing 40 GB of GPU memory.
Inputs and outputs
You provide
Receptor (PDB)
Compound library (SMILES, one per line)
You get back
Library ranked by docking score.
Licence
Commercial use permitted · LGPL-2.1
The LGPL-2.1 licence allows use in commercial research and products.
Questions
Can I use AutoDock-GPU commercially?
Commercial use permitted. The LGPL-2.1 licence allows use in commercial research and products.
How do I run AutoDock-GPU without installing anything?
Use the form under “Run it here” on this page: fill in the inputs and press Run model. It runs on our servers on 4–8× A100/H100, typically in 1.0h, and a run costs about 14,000 credits. Visitors without an account get 1,000 credits a month.
What inputs does AutoDock-GPU accept?
It accepts a receptor structure in PDB format and a small-molecule library provided as SMILES strings, formatted one per line.
What outputs does the tool produce?
It outputs a table of library compounds ranked according to their docking scores.
How is AutoDock-GPU used in multi-stage screening pipelines?
It performs the initial high-throughput screening pass to identify promising binders, whose top poses can then be passed to DiffDock or Boltz-2 for refinement.
Written from AutoDock-GPU's own documentation. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.