Top 40
View all ›BACKFILLED Code and industry signals are partial for backfilled days.
Single-cell
Learned representations turn omics data into interpretable, predictive readouts
Recent preprints and published frameworks couple representation learning with spatial, imaging and single-cell readouts across multiple omics modalities.
Tools & infrastructure
Automated tools for metagenomic binning, fungal ITS abundance, and TE classification
Recent preprints package specific analysis decisions into reproducible tools and repurpose existing human sequencing reads for microbiome profiling.
Genomics
Integrative profiling maps enhancer-promoter wiring and chromatin regulatory logic
Joint single-cell and HiChIP-based profiling now resolves regulatory wiring across layers rather than through isolated single assays.
Structural bio & protein design
Learned and knowledge-based potentials advance protein-ligand affinity and pose scoring
New preprints and papers pair curated PDB complex datasets with graph networks and knowledge-based potentials for binding affinity and pose scoring.
ML / AI methods
Generative and pretrained models learn reusable representations of noisy omics data
Noisy, count-valued, and batch-affected omics measurements are driving adaptation of modern ML architectures to domain-specific data structure.