All Competitors

Inverse folding model refined by online reinforcement learning against folding and stability rewards, cutting design failure rates by 36-48%.

Hydration-site prediction placing water molecules on protein surfaces by score-based diffusion, within 0.3 Å of crystallographic positions.

Immunogenicity classifier for reverse vaccinology, fusing frozen protein language model embeddings with Foldseek and ESM3 structure tokens.

Variable-length generative protein design across structure, sequence, motif scaffolding, and peptide co-design via a generalized Poisson flow.

Protein language model reading each residue alongside an unsupervised local-fragment token, so one encoder serves residue- and chain-level tasks.

Small-molecule hit-discovery pipeline using Boltz-2 co-folding and affinity prediction to rank in-stock compounds or make-on-demand chemical space.

Protein surface tokenizer that turns surface-exposed residues into codebook tokens, lifting SKEMPI binding affinity change prediction to r = 0.600.

Geometric deep learning model that learns atomic-scale representations of molecular interfaces across proteins, small molecules, and nucleic acids.

Flow-matching model for therapeutic peptide design that co-designs sequence, structure, and molecular surface to disrupt protein-protein interactions.

De novo protein binder and nanobody design pipeline that ranks candidates by a protein-protein interaction model rather than structural confidence.

Protein conformation and dynamics generation from MD data, sampling trajectories, independent ensembles, and interpolations between two known states.

Protein question-answering model that fuses sequence and structure into an LLM prompt as virtual tokens, answering free-form questions about function.

Protein multi-conformation predictor that scores per-residue flexibility, then masks MSA columns to steer AlphaFold 2 toward alternative states.

Protein function prediction via compressed in-context learning on a sequence-structure language model, cutting 751-token demonstrations to under 16.

Flow-matching generative model for de novo atomistic protein binder design against protein and small-molecule targets, including carbohydrate binders.

Structure-based drug design model generating 3D ligands inside a protein pocket, aligned by Best-of-K fine-tuning on drug-likeness and docking.

Structure-free ligand generation conditioned on a protein sequence alone, using masked diffusion over SMILES trained on 1.2M BindingDB active pairs.

Protein language model pretrained on structural domain segments, encoding fold and contact signals for remote-homology detection from sequence alone.

Protein conformational ensemble and dynamics generator using latent diffusion to sample all-atom MD trajectories and transition pathways.

Diffusion model that generates 3D small molecules conditioned on protein pockets and partial fragments encoded as continuous spatial density maps.

Unified science foundation model treating molecules, proteins, RNA, DNA, and materials as one sequence language, in 1B, 8B, and 46.7B sizes.

Protein foundation model predicting a ten-modality joint embedding from sequence alone. Reaches 69.99% enzyme class accuracy on dark proteins.

Structure-based mutational effect prediction from local atomic environments, scoring how substitutions change protein stability and binding affinity.