Northwestern University

Large language model trained on functional genomics data to prioritize novel therapeutic targets from genome-wide CRISPR knockout screens.

Protein language model that emulates molecular dynamics, generating equilibrium conformational ensembles and multi-timescale dynamic trajectories.

Histology nuclei segmentation that adapts SAM to train on several datasets at once, aligning auxiliary domains without diluting the primary one.

Autoregressive transformer pretrained on 2,000 hours of intracortical spiking activity, decoding motor intent across subjects, species, and tasks.

4B-parameter generative genome foundation model trained on assembled environmental metagenomes for microbial representation and de novo DNA design.

Structure-to-sequence protein design network inverting trRosetta, predicting coevolution features from a backbone to generate stable sequences.

Glucose forecasting from continuous glucose monitor streams over a two-hour horizon. Cuts one-hour rMSE 48.51% on OhioT1DM without training on it.

Genomic language model that labels adapter sequences in nanopore direct-RNA reads base by base, then splits the chimeric reads those adapters create.

Protein conformational ensemble generation guided by experimental observables, steering a pretrained diffusion sampler toward Boltzmann statistics.

DNA embedding model built on DNABERT-2, using contrastive learning to cluster sequences by species for metagenomic binning without labeled data.

Spot detection for single-molecule RNA FISH and fluorescence microscopy, trained on a differentiable F1 approximation, needing no threshold tuning.

Multi-species genomic foundation model swapping k-mer tokenization for byte pair encoding, matching Nucleotide Transformer with 21x fewer parameters.

Bidirectional transformer for DNA using k-mer tokenization, fine-tunable for promoter, splice site, and transcription factor binding prediction.