Stony Brook University

De novo peptide sequencing transformer that reads modified and unmodified peptides directly from tandem mass spectra without a reference database.

Neuro-oncology foundation model for brain tumor MRI, using distributionally robust pretraining for molecular subtyping and survival prediction.

Generative histopathology foundation model: a diffusion transformer trained on 30M H&E tiles, conditioned on self-supervised slide embeddings.

Enhancer prediction from DNA sequence alone, reaching 88.05% accuracy and 76.22% MCC on held-out ENCODE cCRE regions and annotating the whole genome.

Learned compression autoencoders for histopathology whole-slide images, tuned so reconstructions preserve the features downstream models rely on.

Multi-scale latent diffusion model for histopathology that synthesizes tissue patches at any magnification and composes them into 4096-pixel images.

Histopathology foundation model aligned to spatial transcriptomics by a cross-modal ranking loss, embedding H&E patches without gene input.

AlphaFold fine-tuned via OpenFold on 944 high-resolution MHC-peptide structures, reaching median peptide RMSD of 0.65 Å on held-out complexes.