Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Computational models infer cell fate, perturbation response and spatial gene programs
Recent work pairs single-cell and spatial transcriptomic data with generative, network and deconvolution models that infer dynamics beyond static cell…
ML / AI methods
Models learn cell identity and tissue structure from single-cell and spatial omics
Recent preprints apply transformers and graph models to H&E-based expression prediction, spatial domain identification, and clustering evaluation.
Single-cell
Learned encoders pair cellular context with compound or perturbation representations
Foundation models such as scGPT and UCE are being transferred to drug response and perturbation prediction tasks.
Genomics
Learned models infer copy number and methylation variation from sequencing data
Shallow and long-read sequencing now pair with deep learning and foundation models to extract variation from cfDNA, cfRNA, and spatial data.
Systems biology
Higher-order and topological methods map disease-linked molecular correlation anomalies
Multi-omic and transcriptomic frameworks now aim to convert disease profiles into comparable, interpretable biomarker and pathway readouts.