Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Graph and prior-informed models infer hidden regulatory structure from omics data
Graph, prior-guided and topology-based methods now target hidden regulatory structure and patient subgroups across single-cell, bulk and spatial omics.
Transcriptomics
Spatial transcriptomics methods infer context-resolved cell-cell and gene relationships
Spatial profiles now reach over 400,000 cells, motivating methods that resolve niche-, condition-, and state-dependent expression relationships.
Single-cell
Learned representations fuse single-cell, spatial and bulk omics into latent spaces
Multimodal foundation models and cross-assay harmonization are now being evaluated against defined biological and clinical prediction tasks.
ML / AI methods
Augmenting protein language models with structure, dynamics, and graph-based inputs
Recent work pairs pretrained protein language model embeddings with structural and dynamics data for function and binding prediction.
ML / AI methods
Protein language model embeddings drive sequence-based functional prediction models
Recent models pair protein language model embeddings with attention or graph architectures to predict enzymatic identity, interactions, and binding sites.