RoseTTAFold-AllAtom: RoseTTAFold All-Atom (RFAA) is a deep network capable of modeling full biological assemblies containing proteins, nucleic acids, small molecules, metals, and covalent modifications with high accuracy.
DynamicBind: DynamicBind recovers ligand-specific conformations from unbound protein structures (e.g. AF2-predicted structures), promoting efficient transitions between different equilibrium states.
NeuralPlexer: deep generative model to jointly predict protein-ligand complex 3D structures and beyond.
Allosteric sites
FASTDock: FASTDock is a pipeline for allosteric drug discovery, offering scripts and a Jupyter notebook for efficiently generating and analyzing docking grids, clusters, and fingerprint screenings.
Blind Docking
CBDock2: CBDock2 is an improved protein-ligand blind docking tool integrating cavity detection, docking, and homologous template fitting to suggest novel therapeutic targets for biological and pharmaceutical studies.
CoBDock: CoBDock is a reference implementation of the COBDock algorithm, detailing steps for setup and execution on Linux, with a focus on integrating various molecular docking and pocket identification algorithms.
Classical
AutoDock Vina:
BR-NiB: Brute Force Negative Image-Based Optimization. A docking rescoring method that relyes on shape/ electrostatic potential similarity between the docking poses of ligands and the cavity-based negative images (standalone).
DockThor:
FWAVina: The content for FWAVina was not accessible due to restrictions or an error from the URL provided.
GeauxDock: An ultra-fast automated docking program from LSU, predicting how small ligands bind to macromolecules using a novel hybrid force field and a Monte Carlo protocol.
GLOW-IVES: Provides Python implementation of GLOW (auGmented sampLing with softened vdW potential) and IVES (Iterative Ensemble Sampling) protocols for pose sampling, along with new cross-docking datasets.
HESS:
JAMDA:
labodock: LABODOCK offers a collection of Jupyter Notebooks for molecular docking on Google Colab with minimal coding, streamlining pre- and post-docking processes.
MzDock:
parallel-PLANTS: Offers a method for parallel molecular docking using the PLANTS software, aimed at academic use.
ProBiS-Dock: A Hybrid Multitemplate Homology Flexible Docking Algorithm Enabled by Protein Binding Site Comparison (standalone).
QVINA: QuickVina 2 aims to accurately speed up AutoDock Vina, providing up to 20.49-fold acceleration with high correlation in binding energy prediction.
RDPSO_Vina: A fast docking tool utilizing random drift particle swarm optimization based on the AutoDock Vina and PSOVina framework.
restretto: Effective Protein–Ligand Docking Strategy via Fragment Reuse and a Proof-of-Concept Implementation (standalone).
RxDock: a fast, versatile, and open-source program for docking ligands to proteins and nucleic acids (standalone).
SMINA: A fork of AutoDock Vina that supports scoring function development and high-performance energy minimization, maintained by the University of Pittsburgh.
Uni-Dock: Uni-Dock is a GPU-accelerated molecular docking program that supports various scoring functions and achieves significant speed-up compared with AutoDock Vina on a single CPU core.
VinaCarb: The content for VinaCarb was not available from the URL provided.
VinaGPU2.0:
VinaXB (halogen-bonding): Introduces a halogen bonding scoring function (XBSF) in AutoDock Vina, termed AutoDock VinaXB, to improve docking accuracy with halogenated ligands.
Surflex-Tools: starting with version 4 (standalone).
Consensus
dockECR:
DockingPie: DockingPie is a PyMOL plugin that facilitates consensus docking and scoring analyses, integrating four docking programs (Smina, Autodock Vina, RxDock, and ADFR) to offer a versatile platform for molecular and consensus docking.
Exponential Consensus Ranking:
VoteDock:
Flexible Docking
ADFR: AutoDockFR is a protein-ligand docking program supporting selective receptor flexibility and covalent docking, part of the ADFR suite for streamlined docking procedures.
DSDPFlex:
FlexAID:
GNINA: GNINA is a molecular docking program that incorporates scoring and optimization of ligands using convolutional neural networks, aiming to combine the versatility of smina and AutoDock Vina with the predictive power of deep learning.
iDock: iDock is a multithreaded virtual screening tool for flexible ligand docking in computational drug discovery, inspired by AutoDock Vina and hosted on GitHub under Apache License 2.0.
PackDock: Describes PackDock as a diffusion-based side chain packing model for flexible protein-ligand docking, indicating code will be available following the publication of their paper.
tiny_IFD: Offers lightweight induced fit docking capabilities.
hybrid-SA-IFD: Robust Induced Fit Docking Approach with the Combination ofthe Hybrid All-Atom/United-Atom/Coarse-Grained Model andSimulated Annealing
Fragment-based
Spresso: an ultrafast compound pre-screening method based on compound decomposition
HPC enabled
POAP:
VinaLC:
VinaMPI:
VinaSC:
MD-based
ColDock:
ML-based
ArtiDock:
Re-Dock:
EDM-Dock:
FlexPose:
AQDNet: Implements a Deep Neural Network for Protein-Ligand Docking Simulation, focusing on identifying correct binding poses through convolutional neural network approaches.
CarsiDock: a deep learning paradigm for accurate protein–ligand docking and screening based on large-scale pre-training (standalone, 2024).
DeepBindBC: A practical deep learning method for identifying native-like protein-ligand complexes in virtual screening (online).
DiffBindFR:
DiffDock: A state-of-the-art method for molecular docking, incorporating diffusion steps and a significant improvement in performance and generalization capacity.
DiffDock-Pocket:
ESF - scalar fields: Implements Equivariant Scalar Fields for Molecular Docking with Fast Fourier Transforms, a machine learning-based ligand pose scoring function for rapid optimization.
EViS: EViS is an enhanced virtual screening method integrating ligand docking, protein pocket template searching, and ligand template shape similarity calculations, utilizing a novel PL-score for evaluation.
FABind:
FABind+:
GAABind: GAABind is a Geometry-Aware Attention-Based Network for accurate protein-ligand binding pose and binding affinity prediction, featuring a comprehensive environment setup and dataset processing guide.
GNINA: GNINA is a molecular docking program that incorporates scoring and optimization of ligands using convolutional neural networks, aiming to combine the versatility of smina and AutoDock Vina with the predictive power of deep learning.
GNINA_KD:
gnina-torch: A PyTorch implementation of the GNINA molecular docking scoring function, designed for enhanced performance and adaptability.
Interformer: code to be released, a unified model built upon the Graph-Transformer architecture, which specially crafted to capture non-covalent interactions through the interaction-aware mixture density network
KarmaDock:
PLANTAIN: Predicting LigANd pose wiTh an AI scoring functioN
PointVS: SE(3)-equivariant point cloud networks designed for virtual screening, enabling E(3)-invariant predictions of binding pose and affinity using networks based on the EGNN graph neural network layer.
SurfDock: Surface-Informed Diffusion Generative Model for Reliable and Accurate Protein-ligand Complex Prediction
SurfDock: A Surface-Informed Diffusion Generative Model for reliable and accurate protein-ligand complex prediction, integrating generative model techniques for enhanced docking predictions.
TankBind: Trigonometry-Aware Neural NetworKs for Drug-Protein Binding Structure Prediction
TopoFormer: A topological transformer for protein-ligand complex interaction prediction, integrating multiscale topology techniques with a structure-to-sequence transformer model.
Uni-Mol: A Universal 3D Molecular Representation Learning Framework
Uni-Mol v2:
vScreenML: A machine learning classifier designed for virtual screening, allowing for the rescoring of hits to eliminate false positives, based on the Dataset of Congruent Inhibitors and Decoys (D-COID).
FeatureDock: Protein-Ligand Docking Guided by Physicochemical Feature-Based Local Environment Learning using Transformer
GalaxyDock-DL: Protein–Ligand Docking by Global Optimization and Neural Network Energy
LigPose:
PGBind: pocket-guided explicit attention learning for protein–ligand docking
MetalloProteins
MetalDock:
Multi-Ligand
HARMONICFlow:
Negative Image based
PANTHER: Negative Image based docking and scoring
O-LAP:
Pose Optimisation
DeepRMSD-Vina: DeepRMSD+Vina is a computational framework integrating ligand binding pose optimization and screening, utilizing deep learning alongside the classical Vina scoring function.
Protein Docking
EquiDock: EquiDock employs geometric deep learning for fast and accurate rigid 3D protein-protein docking, focusing on efficiency and accessibility with comprehensive preprocessing and training guidelines.
LightDock: The open-source macromolecular docking framework written in Python
PyDock3: Electrostatics and desolvation for effective scoring of rigid-body protein-protein docking
Quantum
Zhang et al.:
Reviews
Revolutionizing drug discovery: an AI-powered transformation of molecular docking:
Shape-based
PheSA: An Open-Source Tool for Pharmacophore-Enhanced Shape Alignment
Template
FitDock: fits initial conformation to the given template using a hierachical multi-feature alignment approach, subsequently explores the possible conformations, and finally outputs refined docking poses
Water
WatVina: Watvina facilitates drug design with support for explicit or implicit waters, pharmacophore, or position-constrained docking, and external torsion parameters, enhancing the Autodock Vina engine.
WebServers
MolModa: accessible and secure molecular docking in a web browser.
