Université Paris Cité

RNA 3D structure prediction from a single sequence, pairing a frozen language model with base-pair probabilities in a compact folding trunk.

Dense self-supervised pretraining that learns pixel- and voxel-level features for segmenting cardiac MRI and abdominal CT from very few labels.

Anti-phage defense gene predictor fusing protein language model embeddings with a contrastive genomic-context transformer over 64-gene windows.

Template-guided 3D ligand pose generation by flow matching plus differentiable refinement, placing 77.8% of poses within 2 Å RMSD on AlignDockBench.

Protein language model fine-tuned to score any bacterial protein for anti-phage defense function, detecting homology too remote for HMM profiles.

Genomic language model reading bacterial gene neighborhoods as sentences of protein-family tokens to predict anti-phage defense function.

Spike inference from calcium imaging traces, driven by a multistate GCaMP kinetic model that generates the synthetic data its decoders are trained on.

Graph attention foundation model for spatial transcriptomics that assigns spatial domains zero-shot across gene panels, tissues, and technologies.

Missense variant effect predictor that distills GEMME evolutionary scores into a small head over frozen protein language model embeddings.