DiffDock / DiffDock-L
MIT CSAIL (Corso et al.)
Diffusion-based generative model that treats docking as a generative problem over ligand poses. No pre-specified binding pocket needed.
Best For
Blind docking when binding site is unknown; fast generative pose generation
License
Open Source (MIT)
Strengths
- +Blind docking (no pocket required)
- +Generative approach
- +NVIDIA NIM API available
Limitations
- −Underperforms physics-based on cross-docking
- −No binding free energies
- −Requires GPU
R&D Pipeline Coverage
Related Tools
GNINA
Koes Lab (University of Pittsburgh)
AutoDock Vina-based docking engine augmented with a 3D CNN scoring function. Uses Vina for sampling, CNN for scoring and re-ranking.
FlowDock
Morehead, Cheng Lab (University of Missouri)
Geometric flow matching model that maps apo protein structures to bound complexes for multiple ligands simultaneously. Outputs confidence scores and affinity estimates.
AutoDock Vina
The Scripps Research Institute
Classical rigid receptor, flexible ligand docking using empirical and knowledge-based scoring. The most widely used open-source docking tool.
More in Docking & Screening
GNINA
Koes Lab (University of Pittsburgh)
AutoDock Vina-based docking engine augmented with a 3D CNN scoring function. Uses Vina for sampling, CNN for scoring and re-ranking.
FlowDock
Morehead, Cheng Lab (University of Missouri)
Geometric flow matching model that maps apo protein structures to bound complexes for multiple ligands simultaneously. Outputs confidence scores and affinity estimates.
AutoDock Vina
The Scripps Research Institute
Classical rigid receptor, flexible ligand docking using empirical and knowledge-based scoring. The most widely used open-source docking tool.
Stay updated on DiffDock / DiffDock-L
Weekly newsletter covering AI tool releases, benchmarks, and what practitioners actually use.