Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
ML / AI methods
Graph and probabilistic learners decode spatial and single-cell omics architectures
Spatial and single-cell omics methods increasingly pair graph, Gaussian-process and foundation-model learners with uncertainty or biological priors.
Single-cell
Cell-state deconvolution and single-cell benchmarks extend genomic inference
Atlases, deconvolution tools and benchmarks for single-cell and bulk genomics appeared across preprints and journal articles in early July 2026.
ML / AI methods
Protein language models underpin sequence-based prediction of stability and binding sites
Recent work tests whether sequence-only protein language models match structure-informed predictors for mutation stability effects and binding site detection.
Structural bio & protein design
Learned models span structure-aware design, screening, and peptide activity prediction
Protein language models and generative networks are being coupled to structure verification, screening, and peptide classification workflows.
Transcriptomics
Computational methods resolve spatial gene expression to cell and tissue architecture
Models now link spatial transcriptomics to morphology, cell-level resolution, and spatial statistics, extending scRNA-seq frameworks to tissue context.