Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Transcriptomics
Biological priors constrain deep models of gene expression prediction and deconvolution
Ontology, marker, and graph priors are being built into transcriptomic models to aid generalization and interpretability.
ML / AI methods
Protein language model embeddings are probed for function, structure, and binding
Frozen-probe and attention analyses now test what pretrained sequence models encode, alongside contrastive sequence-structure alignment.
Single-cell
Single-cell annotation and spatial tools expand across omics analysis
Recent tools address reference atlas building, LLM-assisted annotation harmonization and reference-missing cell type discovery in single-cell data.
ML / AI methods
Graph and representation learning models expand drug, target, and variant prediction
Recent Bioinformatics papers apply graph, contrastive, and semi-supervised learning to drug, target, and variant prediction tasks.
Structural bio & protein design
Pretrained protein models predict mutational, binding and fold-switching effects
Recent preprints and papers recalibrate and probe protein language, inverse-folding and co-folding models for variant effect prediction.