All Competitors

A graph foundation model for fMRI brain networks, pretrained across 27 datasets with graph and language prompts for zero/few-shot generalization to unseen disorders.

A hierarchical three-stage cascade that generates chemically modified and macrocyclic peptides in HELM notation, supporting de novo design and constrained infilling.

Geometry-first generative framework that reconstructs single-cell spatial coordinates by integrating scRNA-seq with spatial transcriptomics, without cell-type labels.

Multiscale graph neural network for fixed-backbone protein binder sequence design with a contrastive decoding algorithm to improve target selectivity.

A signed heterogeneous graph foundation model pretrained on the SIGMA-KG knowledge graph for zero-shot drug mode-of-action, clinical response, and drug-drug interaction prediction.

A motif-aware graph diffusion transformer for controllable molecular generation that transfers to unseen properties by learning only lightweight task embeddings with the generator frozen.

Modular deep-learning framework for 3D-structure-based RNA sequence design, pairing a direct GNN predictor (SCRU-Seq) and a diffusion model (SCRU-Diff) built on self-contained RNA units.

Billion-parameter single-cell foundation model performing full self-attention across all 28,000 human genes, integrating Gene Ontology priors via GCN for long-range gene context capture in transcriptomics.

A family of SE(3)-equivariant molecular foundation models pretrained on ~70M DFT conformers, reaching state of the art on TDC ADMET and MoleHB property-prediction benchmarks.

All-atom E(3)-equivariant diffusion model that refines RNA structures by resolving steric clashes and completing missing atoms.

Multimodal molecular foundation model fusing SELFIES sequences, 2D graphs, text descriptions, and knowledge-graph embeddings via contrastive pretraining for property prediction.

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

Predicts protein complex stoichiometry from sequence using protein language model embeddings and a graph neural network, exporting AlphaFold3-ready JSON.

Diffusion-based generative model for structure-based peptide inverse folding, pairing a geometric GNN encoder with a Transformer denoiser to design sequences for a target backbone.

Property-driven protein inverse folding: a ProteinMPNN checkpoint aligned via multi-objective preference optimization to improve developability while preserving structural fidelity.

Graph Transformer foundation model integrating 3 million protein pockets and 5 million molecules as E(3)-equivariant graphs for joint protein-ligand geometric representation learning.

A multimodal, domain-constrained foundation model that self-supervises on ~500K PubChem molecules to jointly predict nine thermophysical properties of small molecules.

A knowledge-driven framework that predicts single-cell transcriptomic responses to small molecules, including zero-shot prediction for drugs with no prior perturbation profiles.

A knowledge-graph foundation model that injects cell-type-specific genetic associations into a biomedical knowledge graph to improve drug indication prediction and repurposing.

A polarizable electrostatic machine-learning interatomic potential extending MACE with long-range induction, trained on 100M OMol25 DFT calculations.

Pocket-conditioned generative model that jointly designs enzyme sequences and substrate-binding pockets conditioned on functional priors and substrate structure.

A lightweight graph-convolutional foundation model for spatial transcriptomics that learns spatially coherent, interpretable spot embeddings via masked central-spot prediction.

Contrastive geometric model that unifies structure-based and ligand-based drug design in one checkpoint, enabling zero-shot virtual screening, target fishing, and pocket selection.

A synthesis-constrained generative model that designs synthesizable PROTAC degraders by sampling reaction templates and building blocks, tuned with reinforcement learning.