All Competitors

DNA language model for variant effect prediction across coding and non-coding regions, using whole-genome alignments of 100 vertebrate species.

MSA-free protein structure prediction that replaces multiple sequence alignments with a protein language model pre-trained on billions of sequences.

Multi-modal ophthalmic foundation model for generalist eye AI, spanning fundus imaging and OCT for disease screening, segmentation, and biomarkers.

Structure-aware protein language model pairing amino acid tokens with Foldseek 3Di structural states, outperforming ESM-2 across 10 downstream tasks.

Abdominal CT segmentation model driven by CLIP text embeddings, covering 25 organs and 6 tumor types with zero-shot extension to new categories.

Missense variant pathogenicity predictor built on AlphaFold 2 representations, scoring variants across the human proteome at 0.940 AuROC on ClinVar.

Histopathology foundation models: self-supervised vision transformers pretrained on millions of whole-slide images for tile-level feature extraction.

Self-supervised foundation model for retinal imaging, pretrained on 1.6 million unlabelled fundus and OCT scans to detect ocular and systemic disease.

Vision-language pre-training framework for universal brain MRI diagnosis, learning from imaging-report pairs to cover more than ten brain diseases.

Discrete diffusion model for protein sequence and MSA generation, enabling controllable de novo design directly in sequence space without structure.

fMRI foundation model pretrained with masked autoencoding on roughly 6,700 hours of recordings for clinical prediction and network discovery.

Medical imaging adaptation of the Segment Anything Model, fine-tuned on 4.6M images and 19.7M masks for promptable segmentation across 10 modalities.

Vision-language foundation model for pathology, fine-tuned from CLIP on 208,414 image-text pairs for zero-shot classification and image retrieval.

Perturbation prediction model that forecasts transcriptional responses to multi-gene CRISPR perturbations from scRNA-seq and a gene-gene graph.

Antibody CDR design framework pairing a pretrained antibody language model with a hierarchical graph neural network for one-shot CDR generation.

Radiology foundation model that reads interleaved 2D and 3D scans with text for diagnosis, visual question answering, and report generation.

Open large language models for natural science, fine-tuned on physics, chemistry, and materials science literature with automated instruction tuning.

Transfer learning framework that predicts single-cell gene expression from ~200kb DNA sequences using Enformer embeddings and a lightweight MLP.

Multimodal medical vision-language model for few-shot visual question answering, learning new imaging tasks from in-context examples at inference.

Google's generalist multimodal biomedical AI that encodes clinical text, medical images, and genomics with a single set of weights across 14 tasks.

Bilingual protein language model that translates bidirectionally between amino acid sequences and the 3Di structural alphabet for inverse folding.

Multi-language transformer framework using five pre-trained language models to predict DNA methylation (6mA, 4mC, 5hmC) across species.

Transformer-based single particle tracker for fluorescence microscopy, using multi-hypothesis attention to link particles at low SNR and high density.

Self-supervised histopathology foundation models for H&E whole-slide images. Phikon-v2 is a ViT-L/16 DINOv2 encoder used for biomarker prediction.