All Competitors

Fragment-protein interaction model trained on cellular chemoproteomics, pairing ESM-2 residue embeddings with bilinear attention over ligand atoms.

Structure scoring model for antibody-antigen and protein-ligand complexes, built on one frozen geometry-aware backbone reused across scoring tasks.

Modality-agnostic transformer pretrained by masked abundance reconstruction on 48,837 proteomics profiles reprocessed from 1,397 PRIDE projects.

Biomolecular complex structure prediction without multiple sequence alignment, for protein-protein, protein-ligand and protein-nucleic acid binding.

Protein structure prediction from multiple sequence alignments, trained across MSA depths so one model spans deep alignments and orphan proteins.

Multi-task cellular foundation model predicting drug sensitivity, perturbation expression and drug-protein binding from one pretrained checkpoint.

Transmembrane topology predictor that calls re-entrant regions and interfacial helices, and assigns each protein to one of 17 biological membranes.

Ordered-water prediction for protein structures by flow matching, adding the crystal symmetry mates that coordinate waters at lattice contacts.

Protein perturbation model conditioning substitution-effect prediction on a learned protein-level regime coordinate over 202 million sequences.

Sequence-only dual-encoder contrastive model that ranks whole molecule libraries against a protein target without 3D structures or per-pair scoring.

Protein language model embedding enrichment that injects structural and dynamical signal as a low-energy residual, using sequence alone at inference.

Antifreeze protein classifier over frozen ESM-2 embeddings, trained only on sequences whose antifreeze activity was measured in the lab.

Protein language model that supervises embedding geometry with inter-residue contacts, so representation distance tracks physical distance.

Designs macrocyclic peptide molecular glues bridging two target proteins from sequence alone, validated as VHL-recruiting degraders in cells.

Protein-ligand co-folding model on the OpenFold3 architecture, trained on PDB structures through June 2025 with inference-time chemical steering.

Cryo-EM pose estimation conditioned on a reference volume handed in at inference, assigning particle orientations zero-shot on unseen structures.

Variant effect and disease phosphosite prediction that fuses frozen ESM-2 embeddings with normal-mode protein dynamics over AlphaFold residue graphs.

Binary protein classification over frozen ESM3 embeddings, using transformer contextualization and learned attention pooling instead of fixed pooling.

Long-context co-folding model for protein, nucleic-acid and ligand assemblies, folding systems up to 16,384 residues on a single GPU.

Structure-informed antibody design model pairing a pairformer folding trunk with a preference-trained fitness head for in silico affinity maturation.

Proteochemometric potency model that ranks protein-ligand pairs on the pIC50 scale from a UniProt accession and a SMILES string, no structure needed.

Cross-modal continued pretraining on curated mass-spectrometry proteomes lifts a 70M single-cell model past RNA-only checkpoints far larger.

Biomolecular sequence-structure co-design that plans over frozen folding and inverse-folding models with Monte Carlo tree search, training nothing.