All Competitors

Transformer-based generative language model for de novo RNA design, pretrained on 16 million non-coding RNA sequences from RNAcentral.

Single-cell foundation model built on MAMBA2 state-space blocks, reading all 60,530 genes of a transcriptome without variable-gene selection.

Domain-aware multi-task pretrained 3D Swin Transformer for T1-weighted brain MRI, transferring to Alzheimer's, Parkinson's and brain age tasks.

Generative DNA language model for plasmid design and annotation, pretrained on 153,208 engineered plasmid sequences deposited in Addgene.

Chemical language model that grows molecules one fragment at a time from one or two input fragments, for hit-to-lead design and PROTAC linkers.

Sequence-only allosteric site prediction that reads selected ESM-2 attention heads into a random forest, reaching 0.77 AUPRC without a structure.

De novo protein binder design pipeline that hallucinates binders through frozen AlphaFold2 weights, averaging a 46% experimental success rate.

Equivariant graph neural networks pretrained by denoising, used frozen as geometric featurizers for analyzing conformational dynamics in simulations.

Medical vision-language generalist that answers questions, classifies disease, localizes findings and drafts radiology reports from one checkpoint.

Spatial transcriptomics foundation model using cross-attention over niche ligand genes, pretrained on 4.1M deconvolved human Visium samples.

Generative bit-diffusion model that designs ADAR guide RNAs for any target adenosine, conditioned on desired editing efficiency and specificity.

Conotoxin design framework pairing a ProGen2-small fine-tune on cone snail venom peptides with an autoencoder classifier that screens the samples.

Brain-dynamics foundation model for resting-state fMRI, adapting the Joint-Embedding Predictive Architecture with brain gradient positioning.

Diffusion model for protein-protein docking that unifies pose sampling and energy-based ranking, works without MSAs, and generalizes to new targets.

Multimodal neuroimaging model fusing resting-state fMRI and EEG, pretrained without labels across spatial, temporal, and spectral domains.

Unsupervised VAE-GAN hybrid profiling single-cell morphology in a 10-dimensional disentangled latent space transferable across imaging modalities.

Energy-based diffusion model over protein loop conformations that scores how mutations in peptides and CDR3 loops shift TCR-pMHC binding affinity.

Molecular dynamics surrogate that generates whole trajectories, switching between forward simulation, transition path sampling and upsampling.

Generative protein sequence model built on Bayesian Flow Networks, sampling natural-like de novo sequences and inpainting any subset of residues.

Single-cell foundation model trained on 30 million transcriptomes, binning genes by expression rank to give context-aware gene and cell embeddings.

Cryo-EM and cryo-ET heterogeneity analysis that separates subunit rigid-body motion from compositional change into distinct latent spaces.

Protein-ligand docking network that builds all-atom conformational ensembles of small molecules and side chains from a fixed protein backbone.

Text-guided MRI synthesis model that generates brain MR sequences and resolutions on demand from routine scans using imaging-metadata prompts.