Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
ML / AI methods
Foundation and deep learning models predict transcriptomic, drug and sequence phenotypes
Pretrained tabular, bulk and single-cell foundation models are now benchmarked against specialized predictors across biological scales.
Structural bio & protein design
ML models predict protein binding sites, mutational ΔΔG and stability from structure
Structure-aware deep learning and data-efficient active learning are being applied to binding, stability and design prediction tasks.
Transcriptomics
Graph and generative models reconstruct and integrate spatial transcriptomic data
Seven recent items from July 2026 extend spatial omics methods to imputation, cross-modality alignment, RNA velocity, and 3D segmentation.
ML / AI methods
Learned models predict cancer outcomes, metabolic dependencies and multi-omics structure
Interpretable learning models are now being applied to tumor genomics, metabolic modeling, and spatial multi-omics integration in recent preprints and…
Single-cell
Single-cell multi-omics models infer regulatory networks, copy number and signaling
Recent preprints apply multi-task, graph and change-point methods to paired single-cell multi-omics data for regulatory inference.