All Competitors

De novo drug design model generating target-conditioned ligands by latent diffusion over 1D SELFIES strings, conditioned on protein sequence alone.

Protein-ligand interaction predictor that types seven contact classes between residues and ligand functional groups from sequence and SMILES alone.

Binding free-energy surrogate trained on 1.4M molecular dynamics frames, ranking docking poses ~28,000x faster than physics-based MMGBSA.

Genomic language model predicting drug-induced translational readthrough at premature stop codons, with an R2 of 0.94 across eight compounds.

Structure-based drug design model generating 3D ligands in a protein pocket with interaction-guided flow matching and a learned atom-count predictor.

Structure-based drug design model generating 3D ligands in protein pockets under gradient guidance for affinity, synthesizability, and selectivity.

Protein degrader design framework that mines fragment-target data to build PROTACs and predicts degradation potency (DC50) and maximal degradation.

Energy-based flow matching for 3D molecular structure, using an idempotent predict-and-refine map for protein backbone generation and ligand docking.

Protein-ligand binding model that embeds ligands, pockets, and sequences in hyperbolic space, unifying virtual screening and affinity ranking.

Structure-based virtual screening that rescores docking poses with a deep learning model, reaching 2.6x the enrichment factor of AutoDock Vina.

Structure-based drug design pipeline generating 3D molecules in binding pockets, raising zero-shot CrossDocked2020 docking success from 53% to 64%.

All-atom structure prediction for arbitrary biomolecular complexes of proteins, nucleic acids, and ligands, with code and weights under a BSD license.

Blind protein-ligand docking model adding Ollivier-Ricci curvature descriptors and degree-aware message passing, predicting poses in 0.09 seconds.

Blind flexible protein-ligand docking model trained by two-player self-play, predicting bound ligand and pocket poses in 0.32 seconds per complex.

Generative diffusion transformer for protein-ligand dynamics that produces trajectories, inpaints missing ligand atoms, and samples transition paths.

GPCR ligand bioactivity predictor combining ProteinBERT receptor embeddings with molecular descriptors, spanning the class A receptor family.

Structure-based virtual screening model that jointly predicts protein-ligand complex structures and binding fitness from sequence and SMILES.

Molecular structure elucidation model that reads IR, Raman, UV-Vis, NMR, and mass spectra as text and generates SMILES end to end.

Molecular reasoning language model for molecule captioning and text-to-molecule generation, trained by chain-of-thought distillation then reward RL.

Protein-ligand interaction model pretrained on solvent-aware conformer ensembles, reaching 97.1% AUC on DUD-E virtual screening.

Flow matching model that builds any non-canonical amino acid into a protein pocket from its SMILES string, at 1.43 Å mean RMSD on held-out ncAAs.

Organic reaction foundation model that tokenizes 3D molecular structure to predict products, retrosynthetic routes, conditions, and yields.

Enzyme screening framework pairing a sequence-structure CNN classifier with CLIP-style protein-reaction retrieval to link orphan enzymes to genes.