Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Methods extracting cell-level information from spatial and single-cell transcriptomes
Recent preprints and papers pair new spatial and single-cell tools with calibrated power analysis, matched-modality validation, and interpretable modeling…
ML / AI methods
Pretrained sequence models are probed for structural and functional biology
Pretrained DNA, RNA and protein language models are now judged by verifiable benchmarks and mechanistic probes.
ML / AI methods
Mechanism-structured models make perturbation and molecular predictions interrogable
Single-cell, molecular, and transcriptomic models increasingly embed biological assumptions so their predictions can be tested against mechanism.
Transcriptomics
Learned models transfer and align transcriptomic profiles across samples and modalities
Recent preprints and publications apply domain adaptation, contrastive learning and foundation-model embeddings to recover or interpret transcriptomic profiles.
Structural bio & protein design
Generative molecular models encode geometric, property and synthesis constraints
Diffusion, latent flow and language-model generators now target pocket geometry and synthesis feasibility directly.