Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
ML / AI methods
Multimodal deep learning integrates omics, histology, and spatial data for prediction
Expanding single-cell, spatial, and histology datasets motivate integration methods that handle unaligned sampling and quantify uncertainty.
Structural bio & protein design
Deep models extract directional protein motion from structures and cryo-EM maps
Preprints this period pair graph neural networks and interpretability analysis with MD and cryo-EM data to read directional motion from protein structures.
ML / AI methods
Learned models of drug and perturbation response are being tied to biological structure
Benchmarks and multimodal frameworks now test whether predictive models of drug and perturbation response can yield mechanistically traceable outputs.
Tools & infrastructure
Computational methods integrate single-cell, spatial and genetic multi-omics data
Recent preprints and papers pair scalable joint models and spatial network methods with multiome and spatial atlases of tissue regulation.
Single-cell
Computational methods infer cell dynamics and missing states from single-cell snapshots
Recent preprints and papers recover temporal dynamics, fate bias, and spatial expression from static or sparse single-cell and spatial measurements.