Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Latent generative models predict single-cell perturbation responses across modalities
Flow-matching, Schrödinger bridge, and foundation-model approaches are now applied to perturbation prediction and multi-omics integration in single-cell data.
Systems biology
Graph models learn multi-omic representations for biomarker and patient prediction
Graph-structured and interpretable deep models are being applied to multi-omic, single-cell and clinical data for prediction and mechanistic prioritization.
Structural bio & protein design
Structure-conditioned learned models guide peptide, antibody and ligand design
Recent preprints couple structure prediction and ensemble sampling with generative design, scoring, and experimental or MD verification.
Single-cell
Spatial omics methods couple histology with segmentation and count preprocessing
Spatial platforms now yield dense histology, transcript, and multi-omic data that demand standardized preprocessing, segmentation, and reconstruction.
ML / AI methods
Inference and deep learning methods applied to fMRI, EEG, and MEG data
Preprints and one published study apply simulation-based inference and deep learning to neural recordings, with uncertainty and scaling analyses.