Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Graph-based models of cell-cell interactions and gene regulation in single-cell data
Spatial and multiome single-cell data now support graph models of condition-specific crosstalk, multi-scale interactions, and regulatory networks.
Transcriptomics
Methods target scale bias, longitudinal single-cell dynamics, and regulator inference
Large RNA-seq cohorts and longitudinal single-cell studies demand bias-aware methods for normalization, power estimation, and regulatory inference.
Tools & infrastructure
Learned models infer tumor-cell and spatial expression from histology, MRI, bulk assays
Learned models now predict tumor-cell-specific and spatially resolved quantities from histology, MRI and bulk omics, with external validation.
Structural bio & protein design
Generative models now target multi-objective design of proteins, ligands, and polymers
Recent preprints combine discrete diffusion, reinforcement tuning, and agentic search to optimize several biomolecular properties at once.
ML / AI methods
Biology-informed representations sharpen protein and molecular property prediction
Pretrained protein language models and surface or structure-aware encoders are being adapted with biological priors and confidence signals across tasks.