All Competitors

Blind protein-ligand docking that transfers to binding domains absent from training, scoring 22.6% top-1 on DockGen and 50% on PoseBusters.

All-atom 3D molecular foundation model pretrained across small molecules, proteins, and complexes with an E(3)-equivariant denoising objective.

Compound-protein interaction prediction coupling a chemical language model to a protein language model via a cross-attention block.

Protein-ligand binding affinity prediction from an amino acid sequence and a ligand SMILES string, with no structure, docked pose, or pocket needed.

Discrete graph diffusion transformer generating molecules under several numerical and categorical property targets at once, with no guiding predictor.

Structure-based drug design diffusion model that re-extracts the essential binding subcomplex from a pocket at every step of 3D ligand generation.

Binding energy for protein-ligand, protein-protein, and antibody-antigen complexes is read off an energy model trained on crystal structures alone.

Structure-based RNA virtual screening that scores small molecules against a binding site's base-pairing graph, around 10,000x faster than docking.

Encoder-decoder framework unifying molecules, proteins, and natural language with SELFIES notation for cross-modal drug discovery tasks.

Conditional GAN that generates small molecules against a protein-protein interaction interface, encoding the complex with graph attention networks.

Multimodal biomedical language model that answers free-text questions about molecular graphs and protein sequences as well as literature.

De novo ligand design from a target protein sequence, writing SMILES autoregressively over a merged protein-ligand BPE vocabulary.

Equivariant heterogeneous graph network that rescores docked protein-ligand poses for virtual screening and ranks structural analogs by activity.

Multi-modal LLM answering free-form questions about a compound's indications, pharmacodynamics and mechanism of action from its SMILES string.

Drug pair synergy prediction for rare cancer tissues, read from a language model's representation of a screening row written out as a sentence.

Structure-based drug design by SE(3)-equivariant diffusion over 3D atom coordinates and types, with the same frozen network scoring binding affinity.

Blind protein-ligand docking by diffusion over ligand translations, rotations and torsion angles, reaching 38% top-1 accuracy on PDBBind.

Large-scale chemical language model trained on 1.1 billion SMILES strings using linear attention transformers for molecular property prediction.

Small-molecule generation over SMILES through a fixed-size latent space dense enough that gradient-free search over it optimizes chemical properties.

Shape-based generative model for protein-protein interaction inhibitors, growing new SMILES from the 3D voxel features of a seed compound.

Two SE(3)-equivariant transformers pretrained on 209M molecular conformations and 3M protein pockets, taking 3D coordinates as input and output.

Structure accuracy estimation over atom graphs, scoring macrocyclic peptide and protein-DNA models that residue-level predictors cannot represent.

SMILES language model pretrained on 100M molecules, transferring to forward reaction prediction, retrosynthesis, molecular optimisation, and QSAR.