Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Transcriptomics
Learned models resolve cell states and spatial niches from transcriptomic data
Recent work extends deep generative, graph, and foundation-model methods to multimodal, spatial, and histology-linked transcriptomic data.
Structural bio & protein design
Generative models now design biomolecules under explicit multi-objective constraints
Diffusion, reinforcement fine-tuning, and agentic loops now embed diversity, target-directed objectives, and physics validation into biomolecule generation.
Single-cell
Single-cell dynamics inferred from imaging, splicing, and gene regulatory networks
Single-cell live imaging, morphology, and transcriptomic data are now modeled to infer regulatory structure and cell-state transitions.
ML / AI methods
Interpretability Tools Recover Known Visual Selectivity in Neural Models
Preprints now apply LRP, feature visualization, and mechanistic tools to brain encoders and EEG foundation models, testing whether learned representations…
Tools & infrastructure
Long-read simulators and pipelines are benchmarked against real sequencing data
Nanopore surveillance at continental scale and new read-level benchmarks make simulator and pipeline validation against real reads increasingly urgent.