University of Strasbourg

Modality-agnostic transformer pretrained by masked abundance reconstruction on 48,837 proteomics profiles reprocessed from 1,397 PRIDE projects.

Genomic language model predicting drug-induced translational readthrough at premature stop codons, with an R2 of 0.94 across eight compounds.

CATH superfamily classifier over ProstT5 amino-acid and 3Di structural-alphabet embeddings, reaching 92.2% accuracy on roughly 1,700 superfamilies.

Grounded multimodal language model for endoscopic surgery, supporting visual dialogue, region-based question answering, and bounding-box grounding.

Promptable ultrasound image segmentation foundation model, a SAM adaptation trained on US-43d, the largest public ultrasound segmentation corpus.