All Competitors

Machine learning force field for all-atom protein dynamics, trained on 40 million DFT dipeptide conformations covering backbone and side-chain space.

Gastric pathology model reading whole-slide images through a chain of dependent questions, mirroring a pathologist's stepwise reasoning.

Context-specific protein embeddings across 286 liver disease and cell-type combinations, learned over interactomes built from a single-cell atlas.

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

Virtual spatial transcriptomics model that infers spot-level gene expression from H&E slides by fusing tile, slide, and spatial-position features.

Protein-protein interaction embedding map that infers what a single interaction does from its neighbours among 199,137 human PPIs.

Protein-protein interface prediction that summarizes molecular surface patches with persistent homology descriptors, at 0.77 test AUC.

Medical language model compressed to ternary weights, running a 27B-class clinical and biomedical assistant offline from a single 8.48 GB file.

Antisense oligonucleotide activity prediction from sequence, position-specific chemistry, dose, and cell context. Spearman 0.5970 on ASO Atlas.

Electron microscopy foundation model segmenting mitochondria, ER, nuclei, and lipid droplets across tissues and species from one pretrained encoder.

Open-weights pathology foundation models pairing a 1.1B-parameter ViT-g/8 tile encoder with distilled 86M and 22M variants for H&E histology.

Single-cell foundation model that predicts latent representations of graph-connected gene blocks instead of reconstructing individual gene counts.

Tandem mass spectrum prediction that builds explicit fragmentation pathways, mapping unknown spectra onto 800 million predicted PubChem spectra.

Histopathology foundation model predicting spatial gene expression from H&E slides at single-cell resolution via linear whole-slide attention.

Peptide tandem mass spectrum prediction over a dictionary that pairs every canonical b and y fragment with its integer mass offsets.

Cryo-EM density enhancement for protein-ligand binding sites, sharpening weak ligand maps with a 3D Swin-Conv UNet trained on 6,511 complexes.

Single-cell foundation model for maize, pretrained on a 385,675-cell atlas with Gene Ontology priors for cell typing and cross-species transfer.

Endoscopy vision-language foundation model pretrained on 348K gastrointestinal examinations that pair routine clinical reports with image sets.

Spatial transcriptomics foundation model giving gene-, cell- and neighborhood-scale embeddings zero-shot, plus in-silico gene knockout in tissue.

Breast-specialized multimodal pathology foundation model for core needle biopsy diagnosis, with conformal risk control gating report release.

Chemistry embedding model placing SMILES strings and scientific prose in one vector space, top-tier on 43 of 48 molecular and text benchmarks.

Renal tumor histopathology model that detects tissue regions, classifies nine subtypes, grades nuclei and scores prognosis from a single H&E slide.

Diffusion super-resolution for H&E whole-slide images, steering denoising with SAM-derived structural anchors and adapted DINOv3 semantic control.