Discover, rate, and verify biology preprints.

This paper develops an AI model to estimate immune biomarkers from H&E images, showing it can predict prognostic outcomes as effectively as costly mIF assays.

REMAG improves eukaryotic genome recovery from metagenomic data by using non-contrastive self-supervised learning, outperforming existing tools in producing near-complete genomes.

The paper introduces a Gaussian random field model for imputing missing data in longitudinal microbiome studies, achieving comparable accuracy to baselines while modeling temporal and multivariate correlations.

EcoNet improves immunotherapy response prediction by modeling tumor microenvironment signaling networks, outperforming existing methods in AUC and stratifying ccRCC patients by clinical outcome.

The paper introduces genomeRxiv, a web server that enables precise strain-level identification of prokaryotes using ANI-based LINs, overcoming limitations of traditional taxonomic classification.

The paper enhances AOP networks by annotating Key Events with Gene Ontology terms, improving transcriptomics-based assessment of KE perturbation in pesticide-induced hepatotoxicity.

The paper extends GTalign to support nucleic acid structures, improving alignment accuracy and reducing runtime for large-scale 3D structure comparisons.

This paper introduces SafeDelta, a split-control criterion for few-shot adaptation in single-cell perturbation prediction, showing it improves reproducibility and identifies biological programs.

The paper evaluates whether single-cell RNA-seq improves the prioritization of antibody-based oncology targets by combining tumor specificity, normal safety, and blood sink features.

This paper develops Scratch-AID+, an automated mouse scratching detection framework, showing that fine-tuned video foundation models outperform classic models in generalizability and accuracy.

The paper introduces RAIDS, a toolset for inferring genetic ancestry from various sequence data, including cancer and low-coverage scenarios.

The paper quantifies hubness in T cell receptor distance spaces and evaluates local scaling as a correction, showing improvements in retrieval accuracy but no universal solution.

BioMod introduces biological modules as intermediate units for single-cell foundation modeling, achieving state-of-the-art performance in cell-type annotation and drug-response prediction.

This study examines the low adoption of computational gene-panel-selection tools in targeted spatial transcriptomics, finding that most researchers still use manual or vendor panels.

SQUALL integrates histology and spatial transcriptomics to enable virtual spatial gene expression profiling and clinical prediction across cancers and platforms.

PanTEon provides a cross-kingdom framework for reproducible TE classification, benchmarking machine learning methods and highlighting performance variations across TE superfamilies and lineages.

PRICE 2 improves detection of short and overlapping ORFs using group LASSO regularization, outperforming existing tools and enabling discovery of more cryptic MHC-I peptides.

The paper evaluates whether label-free morphology can replace targeted cellular measurements by analyzing paired phase-reporter data across gene knockouts and screens.

VirtualSkin predicts intervention responses in skin cells by integrating transcriptomic data with biological descriptions across multiple datasets.