💊📝 Awesome Protein-Ligand Interactions

This is the official repository for the review article "Modeling Protein–Ligand Interactions for Drug Discovery in the Era of Deep Learning" published in Chemical Society Reviews. [link]. We curate papers, tools, and resources related to deep learning-based modeling of protein–ligand interactions for drug discovery.

Note: This repository is not intended to be an exhaustive or definitive collection. The included papers and resources were selected based on the authors’ perspectives and interests, and we sincerely apologize for any important work that may have been unintentionally omitted. We warmly welcome community contributions and suggestions to help improve and expand this resource.

Table of Contents

• Table of Contents
• 1. Deep Learning-Augmented Molecular Dynamics
  • 1.1 MD simulations of protein-ligand complexes with MLFFs
  • 1.2 Learning protein-ligand interactions from MD trajectories
  • 1.3 Generative modeling of MD trajectories for protein-ligand complexes
  • 1.4 Benchmarks, datasets, and tools
• 2. Deep Learning-Enhanced Molecular Docking and Virtual Screening
  • 2.1 Deep learning-based docking methods
  • 2.2 Deep learning scoring functions and binding affinity prediction
  • 2.3 Deep learning-accelerated virtual screening
  • 2.4 Benchmarks, datasets, and tools
• 3. End-to-End Structural Modeling
  • 3.1 Protein structure prediction with applications in drug discovery
  • 3.2 Generative modeling of protein–ligand complexes and equilibrium ensembles
• 4. Structure-Based De Novo Drug Design
  • 4.1 Structure-based de novo drug design methods
  • 4.2 Ligand-based design and lead optimization methods
  • 4.3 Benchmarks, datasets, and tools
• 5. Sequence-Based Methods for Drug Discovery

1. Deep Learning-Augmented Molecular Dynamics

1.1 MD simulations of protein-ligand complexes with MLFFs

Name	Paper Title	Year	Venue	Resources	Notes
/	Simulating protein-ligand binding with neural network potentials [paper]	2020	Chem. Sci.	/	NNP/MM
qmlify	Towards chemical accuracy for alchemical free energy calculations with hybrid physics-based machine learning/molecular mechanics potentials [preprint]	2020	bioRxiv	[code]	NNP/MM
/	NNP/MM: accelerating molecular dynamics simulations with machine learning potentials and molecular mechanics [paper]	2023	JCIM	[code]	NNP/MM
/	Enhancing protein-ligand binding affinity predictions using neural network potentials [paper]	2024	JCIM	[data]	NNP/MM
AP-Net	A physics-aware neural network for protein-ligand interactions with quantum chemical accuracy [paper]	2024	Chem. Sci.	[code]	Native MLFF
espaloma-0.3	Machine-learned molecular mechanics force fields from large-scale quantum chemical data [paper]	2024	Chem. Sci.	[code]	Native MLFF
QuantumBind-RBFE	QuantumBind-RBFE: accurate relative binding free energy calculations using neural network potentials [paper]	2025	JCIM	[code]	NNP/MM

1.2 Learning protein-ligand interactions from MD trajectories

Name	Paper Title	Year	Venue	Resources
NRI-MD	Neural relational inference to learn long-range allosteric interactions in proteins from molecular dynamics simulations [paper]	2022	Nat. Commun.	[code]
ProtMD	Pre-training of equivariant graph matching networks with conformation flexibility for drug binding [paper]	2022	Adv. Sci.	[code]
Dynaformer	From static to dynamic structures: improving binding affinity prediction with graph-based deep learning [paper]	2024	Adv. Sci.	[code]
MDbind	Spatio-temporal learning from molecular dynamics simulations for protein–ligand binding affinity prediction [paper]	2025	Bioinformatics	[code]

1.3 Generative modeling of MD trajectories for protein-ligand complexes

Name	Paper Title	Year	Venue	Resources
BioMD	BioMD: all-atom generative model for biomolecular dynamics simulation [paper]	2026	ICLR	/
BioKinema	Physically grounded generative modeling of all-atom biomolecular dynamics [preprint]	2026	bioRxiv	[code]

1.4 Benchmarks, datasets, and tools

Name	Paper Title	Year	Venue	Resources	Notes
SPICE	SPICE, a dataset of drug-like molecules and peptides for training machine learning potentials [paper]	2023	Sci. Data	[data] [code]	QM calculations
MISATO	MISATO: machine learning dataset of protein-ligand complexes for structure-based drug discovery [paper]	2024	Nat. Comput. Sci.	[data] [code]	QM calculations & MD trajectories
PLAS-20k	PLAS-20k: extended dataset of protein-ligand affinities from MD simulations for machine learning applications [paper]	2024	Sci. Data	[data]	MD trajectories & MMPBSA calculations
OMol25	The open molecules 2025 (OMol25) dataset, evaluations, and models [preprint]	2025	arXiv	[data] [code] [blog]	QM calculations
DD-13M	A Novel 4-D Dataset Paradigm for Studying Complete Ligand-Protein Dissociation Dynamics [preprint]	2025	arXiv	[data] [webpage]	MD trajectories
qcMol	qcMol: a large-scale dataset of 1.2 million molecules with high-quality quantum chemical annotations for molecular representation learning [preprint]	2025	bioRxiv	[webpage]	QM calculations
BMS25	Biomolecular multiscale simulation (BMS25) dataset to train neural network potentials for QM/MM settings with electrostatic embedding [preprint]	2025	ChemRxiv	[data]	QM/MM calculations
AnewFEP	Physics-Based vs AI-Based Free Energy Prediction for Protein-Ligand Potency: Public Benchmarks and Internal Project Evidence [preprint]	2026	ChemRxiv	/	RBFE benchmark & prospective evidence

2. Deep Learning-Enhanced Molecular Docking and Virtual Screening

2.1 Deep learning-based docking methods

Name	Paper Title	Year	Venue	Resources	Notes
DeepDock	A geometric deep learning approach to predict binding conformations of bioactive molecules [paper]	2021	Nat. Mach. Intell.	[code]	Rigid receptor
TankBind	TankBind: trigonometry-aware neural networks for drug-protein binding structure prediction [paper]	2022	NeurIPS	[code]	Rigid receptor
EquiBind	EquiBind: geometric deep learning for drug binding structure prediction [paper]	2022	ICML	[code]	Rigid receptor
E3Bind	E3Bind: an end-to-end equivariant network for protein-ligand docking [paper]	2022	ICLR	/	Rigid receptor
DiffDock	DiffDock: diffusion steps, twists, and turns for molecular docking [paper]	2022	ICLR	[code]	Rigid receptor
Uni-Mol	Uni-Mol: a universal 3D molecular representation learning framework [paper]	2023	ICLR	[code]	Rigid receptor
KarmaDock	Efficient and accurate large library ligand docking with KarmaDock [paper]	2023	Nat. Comput. Sci.	[code]	Rigid receptor
FABind	FABind: fast and accurate protein-ligand binding [paper]	2023	NeurIPS	[code]	Rigid receptor
FlexPose	Equivariant flexible modeling of the protein-ligand binding pose with geometric deep learning [paper]	2023	JCTC	[code]	Side-chain flexible
CarsiDock	CarsiDock: a deep learning paradigm for accurate protein-ligand docking and screening based on large-scale pre-training [paper]	2024	Chem. Sci.	[code]	Rigid receptor
DeltaDock	DeltaDock: a unified framework for accurate, efficient, and physically reliable molecular docking [paper]	2024	NeurIPS	[code]	Rigid receptor
DiffBindFR	DiffBindFR: an SE (3) equivariant network for flexible protein-ligand docking [paper]	2024	Chem. Sci.	[code]	Side-chain flexible
DynamicBind	DynamicBind: predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model [paper]	2024	Nat. Commun.	[code]	Fully flexible
PackDock	Flexible protein–ligand docking with diffusion-based side-chain packing [paper]	2025	PNAS	[code]	Side-chain flexible
Matcha	Matcha: Multi-Stage Riemannian Flow Matching for Accurate and Physically Valid Molecular Docking [preprint]	2025	arXiv	[code]	Rigid receptor

2.2 Deep learning scoring functions and binding affinity prediction

Name	Paper Title	Year	Venue	Resources	Notes
OnionNet	OnionNet: a multiple-layer intermolecular-contact-based convolutional neural network for protein-ligand binding affinity prediction [paper]	2019	ACS Omega	[code]	/
RTMScore	Boosting protein-ligand binding pose prediction and virtual screening based on residue-atom distance likelihood potential and graph transformer [paper]	2022	JMC	[code]	/
PIGNet	PIGNet: a physics-informed deep learning model toward generalized drug-target interaction predictions [paper]	2022	Chem. Sci.	[code]	Physical terms
GenScore	A generalized protein-ligand scoring framework with balanced scoring, docking, ranking and screening powers [paper]	2023	Chem. Sci.	[code]	/
PBCNet	Computing the relative binding affinity of ligands based on a pairwise binding comparison network [paper]	2023	Nat. Comput. Sci.	[code]	Relative binding affinity
PLANET	PLANET: a multi-objective graph neural network model for protein-ligand binding affinity prediction [paper]	2023	JCIM	[code]	Binding pose-free
EquiScore	Generic protein-ligand interaction scoring by integrating physical prior knowledge and data augmentation modelling [paper]	2024	Nat. Mach. Intell.	[code]	Interaction fingerprints
DeepRLI	DeepRLI: a multi-objective framework for universal protein-ligand interaction prediction [paper]	2025	Digit. Discov.	[code]	/
PBCNet2.0	Atomic-level protein–ligand recognition with PBCNet2.0 for probe discovery [paper]	2026	Nat. Chem. Biol.	[code]	Relative binding affinity, probe discovery
BioScore	BioScore: a foundational scoring function for diverse biomolecular complexes [preprint]	2025	arXiv	/	A unified scoring function

2.3 Deep learning-accelerated virtual screening

Name	Paper Title	Year	Venue	Resources	Notes
Deep Docking	Deep docking: a deep learning platform for augmentation of structure based drug discovery [paper]	2020	ACS Cent. Sci.	[code]	/
DrugCLIP	DrugCLIP: contrastive protein-molecule representation learning for virtual screening [paper]	2023	NeurIPS	[code] [project page]	CLIP architecture
OpenVS	An artificial intelligence accelerated virtual screening platform for drug discovery [paper]	2024	Nat. Commun.	[code]	Active learning
/	Rapid traversal of vast chemical space using machine learning-guided docking screens [paper]	2025	Nat. Comput. Sci.	[code]	/
FragmentScope	FragmentScope - exploring the fragment space with learned surface representations [preprint]	2025	bioRxiv	[code]	Surface-based fragment screening
BoltzMol-1	BoltzMol-1: Towards reliable virtual screening for fast and cost-effective hit discovery [technical report]	2026	Technical Report	[platform]	AI-driven prospective hit discovery
CombiDOCK & MINT-Dock	Combinatorial docking and molecular generation to navigate over 100-billion molecules for prospective ligand discovery [preprint]	2026	bioRxiv	/	100B-scale make-on-demand screening

2.4 Benchmarks, datasets, and tools

Name	Paper Title	Year	Venue	Resources	Notes
BindingDB	BindingDB: a web-accessible database of experimentally determined protein-ligand binding affinities [paper]	2007	NAR	[webpage]	Database
ChEMBL	ChEMBL: a large-scale bioactivity database for drug discovery [paper]	2012	NAR	[webpage] [code]	Database
ZINC Database	ZINC 15-ligand discovery for everyone [paper]	2015	JCIM	[webpage]	Database
Enamine REAL Space	Generating multibillion chemical space of readily accessible screening compounds [paper]	2020	iScience	[webpage]	Make-on-demand database
PoseBusters	PoseBusters: AI-based docking methods fail to generate physically valid poses or generalise to novel sequences [paper]	2024	Chem. Sci.	[code] [doc]	Benchmark, pose quality check
Leak Proof PDBBind	Leak Proof PDBBind: A Reorganized Data Set of Protein-Ligand Complexes for More Generalizable Binding Affinity Prediction [paper]	2026	J. Phys. Chem. B	[code]	Dataset split
SPECTRA	Evaluating generalizability of artificial intelligence models for molecular datasets [paper]	2024	Nat. Mach. Intell.	[code]	Framework for model evaluation
SAIR	SAIR: Enabling Deep Learning for Protein-Ligand Interactions with a Synthetic Structural Dataset [paper]	2026	ICLR	[data]	Database, AF3-predicted structures
QUID	Extending quantum-mechanical benchmark accuracy to biological ligand-pocket interactions [paper]	2025	Nat. Commun.	[data]	Benchmark, QM-level pocket-ligand interaction systems
BindFlow	BindFlow: A Free, User-Friendly Pipeline for Absolute Binding Free Energy Calculations Using Free Energy Perturbation or MM(PB/GB)SA [paper]	2026	JCTC	[code]	Tool, FEP & MM(PB/GB)SA pipeline

3. End-to-End Structural Modeling

3.1 Protein structure prediction with applications in drug discovery

Paper Title	Year	Venue	Resources	Notes
AlphaFold2 structures guide prospective ligand discovery [paper]	2024	Science	/	GPCR targets
AlphaFold accelerates artificial intelligence powered drug discovery: efficient discovery of a novel CDK20 small molecule inhibitor [paper]	2023	Chem. Sci.	[PandaOmics] [Chemistry42]	CDK20
AlphaFold accelerated discovery of psychotropic agonists targeting the trace amine-associated receptor 1 [paper]	2024	Sci. Adv.	/	TAAR1
Virtual library docking for cannabinoid-1 receptor agonists with reduced side effects [paper]	2025	Nat. Commun.	[data]	Cannabinoid-1 receptor

3.2 Generative modeling of protein–ligand complexes and equilibrium ensembles

Name	Paper Title	Year	Venue	Resources	Notes
NeuralPLexer	State-specific protein-ligand complex structure prediction with a multiscale deep generative model [paper]	2024	Nat. Mach. Intell.	[code]	/
RoseTTAFold All-Atom	Generalized biomolecular modeling and design with RoseTTAFold All-Atom [paper]	2024	Science	[code]	/
AlphaFold3	Accurate structure prediction of biomolecular interactions with AlphaFold 3 [paper]	2024	Nature	[code] [server]	Non-commercial usage
Umol	Structure prediction of protein-ligand complexes from sequence information with Umol [paper]	2024	Nat. Commun.	[code]	/
Chai-1	Chai-1: decoding the molecular interactions of life [preprint]	2024	bioRxiv	[code] [server]	/
Protenix	Protenix - Advancing Structure Prediction Through a Comprehensive AlphaFold3 Reproduction [preprint]	2025	bioRxiv	[code] [server]	/
Boltz-1	Boltz-1: democratizing biomolecular interaction modeling [preprint]	2024	bioRxiv	[code] [blog]	/
Boltz-2	Boltz-2: Towards accurate and efficient binding affinity prediction [preprint]	2025	bioRxiv	[code] [design-code]	FEP-level affinity prediciton
GeoFlow-V2	GeoFlow-V2: a unified atomic diffusion model for protein structure prediction and de novo design [preprint]	2025	bioRxiv	[server]	Protein Binder & antibody design
Chai-2	Zero-shot antibody design in a 24-well plate [preprint]	2025	bioRxiv	[blog]	Minibinder & antibody design
RF3	Accelerating biomolecular modeling with AtomWorks and RF3 [preprint]	2025	bioRxiv	[code]	AtomWorks framework
Pearl	Pearl: a foundation model for placing every atom in the right location [preprint]	2025	arXiv	/	Synthetic data, SO(3)-equivariance
SeedFold	SeedFold: scaling biomolecular structure prediction [preprint]	2025	arXiv	[project page]	Synthetic data, linear triangular attention
Protenix-v1	Protenix-v1: toward high-accuracy open-source biomolecular structure prediction [preprint]	2026	bioRxiv	[code] [server]	Open-source structure prediction model with superior performance to AlphaFold3
Protenix-v2	Protenix-v2: Broadening the Reach of Structure Prediction and Biomolecular Design [preprint]	2026	bioRxiv	[code] [server]	Structure prediction and biomolecular design
IsoDDE	Accurate predictions of novel biomolecular interactions with IsoDDE [technical report]	2026	Technical Report	/	State-of-the-art performance in challenging structure modeling and affinity prediction, implementation details not disclosed
OpenFold3	OpenFold3-preview2 technical report [technical report]	2026	Technical Report	[code]	MGnify 13M-sequence distillation dataset openly available
AnewSampling	Learning the all-atom equilibrium distribution of biomolecular interactions at scale [preprint]	2026	bioRxiv	[project page]	Dynamic equilibrium sampling of protein–ligand complexes

4. Structure-Based De Novo Drug Design with Deep Generative Models

4.1 Structure-based de novo drug design methods

Name	Paper Title	Year	Venue	Resources	Notes
/	A 3D generative model for structure-based drug design [paper]	2021	NeurIPS	[code]	Autoregressive
DeepLigBuilder	Structure-based de novo drug design using 3D deep generative models [paper]	2021	Chem. Sci.	[data]	Autoregressive
GraphBP	Generating 3D molecules for target protein binding [paper]	2022	ICML	[code]	Autoregressive
Pocket2Mol	Pocket2Mol: efficient molecular sampling based on 3D protein pockets [paper]	2022	ICML	[code]	Autoregressive
DeepLigBuilder+	Synthesis-driven design of 3D molecules for structure-based drug discovery using geometric transformers [preprint]	2022	arXiv	/	Autoregressive, synthon, pharmacophore
TargetDiff	3D equivariant diffusion for target-aware molecule generation and affinity prediction [paper]	2023	ICLR	[code]	Non-autoregressive
FLAG	Molecule generation for target protein binding with structural motifs [paper]	2023	ICLR	[code]	Autoregressive, fragment
ResGen	ResGen is a pocket-aware 3D molecular generation model based on parallel multiscale modelling [paper]	2023	Nat. Mach. Intell.	[code]	Autoregressive
DrugGPS	Learning subpocket prototypes for generalizable structure-based drug design [paper]	2023	ICML	[code]	Autoregressive, fragment
DecompDiff	DecompDiff: diffusion models with decomposed priors for structure-based drug design [paper]	2023	ICML	[code]	Non-autoregressive, fragment
D3FG	Functional-group-based diffusion for pocket-specific molecule generation and elaboration [paper]	2023	NeurIPS	[code]	Non-autoregressive, fragment
SurfGen	Learning on topological surface and geometric structure for 3D molecular generation [paper]	2023	Nat. Comput. Sci.	[code]	Autoregressive, explicit pocket surface features
PocketFlow	PocketFlow is a data-and-knowledge-driven structure-based molecular generative model [paper]	2024	Nat. Mach. Intell.	[code]	Autoregressive
DiffSBDD	Structure-based drug design with equivariant diffusion models [paper]	2024	Nat. Comput. Sci.	[code]	Non-autoregressive
PMDM	A dual diffusion model enables 3D molecule generation and lead optimization based on target pockets [paper]	2024	Nat. Commun.	[code]	Non-autoregressive
MolCRAFT	MolCRAFT: structure-based drug design in continuous parameter space [paper]	2024	ICML	[code]	Non-autoregressive
Lingo3DMol	Generation of 3D molecules in pockets via a language model [paper]	2024	Nat. Mach. Intell.	[code]	Autoregressive LM, fragment
KGDiff	KGDiff: towards explainable target-aware molecule generation with knowledge guidance [paper]	2024	Brief. Bioinform.	[code]	Non-autoregressive, Vina scoring guidance
FlexSBDD	FlexSBDD: structure-based drug design with flexible protein modeling [paper]	2024	NeurIPS	/	Non-autoregressive, dynamic pocket
DynamicFlow	Integrating protein dynamics into structure-based drug design via full-atom stochastic flows [paper]	2025	ICLR	/	Non-autoregressive, dynamic pocket
DrugFlow & FlexFlow	Multi-domain distribution learning for de novo drug design [paper]	2025	ICLR	[code]	Non-autoregressive, flexible side-chains
RxnFlow	Generative flows on synthetic pathway for drug design [paper]	2025	ICLR	[code]	Autoregressive, codesigns synthetic routes
SynGFN	SynGFN: learning across chemical space with generative flow-based molecular discovery [paper]	2025	Nat. Comput. Sci.	[code]	Autoregressive, codesigns synthetic routes
PocketXMol	Unified modeling of 3D molecular generation via atomic interactions with PocketXMol [paper]	2026	Cell	[code]	Non-autoregressive, atom-level generative foundation model
AnewOmni	Programming biomolecular interactions with all-atom generative model [preprint]	2026	bioRxiv	[project page]	Non-autoregressive, all-atom latent diffusion, unified generation of peptides, antibodies, and small molecules

4.2 Ligand-based design and lead optimization methods

Name	Paper Title	Year	Venue	Resources	Notes
Delinker	Deep generative models for 3D linker design [paper]	2020	JCIM	[code]	Linker
DEVELOP	Deep generative design with 3D pharmacophoric constraints [paper]	2021	Chem. Sci.	[code]	Linker, pharmacophore
DeepHop	Deep scaffold hopping with multimodal transformer neural networks [paper]	2021	J. Cheminform.	[code]	Scaffold hopping
DRLinker	DRLinker: deep reinforcement learning for optimization in fragment linking design [paper]	2022	JCIM	[code]	Linker
SILVR	SILVR: guided diffusion for molecule generation [paper]	2023	JCIM	[code]	Training-free
FFLOM	FFLOM: a flow-based autoregressive model for fragment-to-lead optimization [paper]	2023	JMC	[code]	Linker
LinkerNet	LinkerNet: fragment poses and linker co-design with 3D equivariant diffusion [paper]	2023	NeurIPS	[code]	Linker
PGMG	A pharmacophore-guided deep learning approach for bioactive molecular generation [paper]	2023	Nat. Commun.	[code]	Scaffold hopping, pharmacophore
DiffLinker	Equivariant 3D-conditional diffusion model for molecular linker design [paper]	2024	Nat. Mach. Intell.	[code]	Linker
ShEPhERD	ShEPhERD: diffusing shape, electrostatics, and pharmacophores for bioisosteric drug design [paper]	2025	ICLR	[code]	Bioisosteric lignd design, shape & electrostatics & pharmacophore
Delete	Deep lead optimization enveloped in protein pocket and its application in designing potent and selective ligands targeting LTK protein [paper]	2025	Nat. Mach. Intell.	[code]	/
TransPharmer	Accelerating discovery of bioactive ligands with pharmacophore-informed generative models [paper]	2025	Nat. Commun.	[code]	Scaffold elaboration, pharmacophore
PhoreGen	Pharmacophore-oriented 3D molecular generation toward efficient feature-customized drug discovery [paper]	2025	Nat. Comput. Sci.	[code]	Pharmacophore
ED2Mol	Electron-density-informed effective and reliable de novo molecular design and optimization with ED2Mol [paper]	2025	Nat. Mach. Intell.	[code]	Electron density

4.3 Benchmarks, datasets, and tools

Name	Paper Title	Year	Venue	Resources	Notes
GuacaMol	GuacaMol: benchmarking models for de novo molecular design [paper]	2019	JCIM	[code]	Benchmark
MOSES	Molecular sets (MOSES): a benchmarking platform for molecular generation models [paper]	2020	Front. Pharmacol.	[code]	Benchmark
CrossDocked2020	Three-dimensional convolutional neural networks and a cross-docked data set for structure-based drug design [paper]	2020	JCIM	[data] [instruction]	Benchmark, dataset
POKMOL-3D	How good are current pocket-based 3D generative models?: the benchmark set and evaluation of protein pocket-based 3D molecular generative models [paper]	2024	JCIM	[code]	Benchmark
Durian	Durian: a comprehensive benchmark for structure-based 3D molecular generation [paper]	2024	JCIM	[code]	Benchmark
CBGBench	CBGBench: fill in the blank of protein-molecule complex binding graph [paper]	2025	ICLR	[code]	Benchmark
MolGenBench	Benchmarking real-world applicability of molecular generative models from de novo design to lead optimization with MolGenBench [preprint]	2025	bioRxiv	[code]	Benchmark
TarPass	Revisiting target-aware de novo molecular generation with TarPass: between rational design and Texas Sharpshooter [paper]	2026	Advanced Science	[code]	Benchmark

5. Sequence-Based Methods for Drug Discovery

Name	Paper Title	Year	Venue	Resources	Notes
DrugBAN	Interpretable bilinear attention network with domain adaptation improves drug--target prediction [paper]	2023	Nat. Mach. Intell.	[code]	DTI prediction
DeepTarget	Deep generative model for drug design from protein target sequence [paper]	2023	J. Cheminform.	[code]	Target-conditioned generation
ConPLex	Contrastive learning in protein language space predicts interactions between drugs and protein targets [paper]	2023	PNAS	[code]	ULVS
TransformerCPI 2.0	Sequence-based drug design as a concept in computational drug design [paper]	2023	Nat. Commun.	[code]	DTI prediction
CogMol	Accelerating drug target inhibitor discovery with a deep generative foundation model [paper]	2023	Sci. Adv.	[code]	Target-conditioned generation
AI-Bind	Improving the generalizability of protein-ligand binding predictions with AI-Bind [paper]	2023	Nat. Commun.	[code]	DTI prediction, interaction network
DRAGONFLY	Prospective de novo drug design with deep interactome learning [paper]	2024	Nat. Commun.	[code]	DTI prediction, generation, interaction network
PSICHIC	Physicochemical graph neural network for learning protein-ligand interaction fingerprints from sequence data [paper]	2024	Nat. Mach. Intell.	[code]	Interaction fingerprints
DeepBlock	A deep learning approach for rational ligand generation with toxicity control via reactive building blocks [paper]	2024	Nat. Comput. Sci.	[code]	Target-conditioned generation, fragment
LaMGen	LaMGen: LLM-based 3D molecular generation for multi-target drug design [paper]	2026	Nat. Commun.	[code] [software]	Multi-target 3D generation from protein sequences

Contributing

We welcome contributions from the community. To contribute, please fork this repository, apply your modifications, and open a pull request to the main branch.

License

This repository is licensed under the MIT License. See the LICENSE file for details.

Citation

If you find this repository useful for your research, please consider citing the following paper:

@article{wang2025modeling,
  title={Modeling protein–ligand interactions for drug discovery in the era of deep learning},
  author={Wang, Yuzhe and Li, Yibo and Chen, Jiaxiao and Lai, Luhua},
  journal={Chemical Society Reviews},
  year={2025},
  publisher={The Royal Society of Chemistry},
  doi={10.1039/D5CS00415B}
}