Discover, rate, and discuss biology preprints.

The paper benchmarks sequence- and structure-based protein models using a unified framework, showing structure-based models outperform others in generating functional variants.

This paper introduces Bone-Segnet, a dual-view guided network for automated bone lesion segmentation in SPECT images, achieving improved Dice scores and enabling quantitative analysis of lesion characteristics.

TD-COM detects personalized early-warning signals for critical transitions in complex diseases using single-sample community detection from transcriptomic data.

ConvergeCELL translates patient transcriptomics into therapeutic hypotheses by integrating patient representation, interpretability, and large language models, outperforming baselines in gene recovery and classification.

This paper shows that biologically informed neural networks (BINNs) cannot reliably recover internal biological structures from outputs alone due to nonidentifiability.

CellNeighborEX v2 identifies context-specific cell-cell interaction genes from spatial transcriptomics data without relying on ligand-receptor databases.

MurineCyto-Det provides a high-resolution, annotated murine BALF cytology dataset for cell segmentation and detection, highlighting challenges in automated analysis.

WasteFams catalogs 3,887 non-redundant protein families from wastewater metagenomes and metatranscriptomes, providing structural, taxonomic, and functional annotations for microbial function discovery.

BindPred predicts protein-protein binding affinity from amino acid sequences using language model embeddings and gradient boosting trees, achieving high correlation on a benchmark dataset.

PARiS identifies and removes technical isomiR errors in sRNA-seq data, improving miRNA-level differential expression analysis.

WormSpot is a machine learning tool for scoring C. elegans viability in Candida infections, achieving accurate results with minimal image input and efficient validation.

The paper identifies early plasma proteomic changes preceding amyloidosis diagnosis, showing cardiac and immune dysregulation before clinical symptoms.

The paper introduces CLIC scores to quantify confidence in gene-peak associations for single-cell RNA-ATAC integration, improving integration accuracy and biological interpretability.

ExposoGraph is an interactive knowledge-graph platform that visualizes carcinogen metabolism and DNA damage pathways, integrating data from multiple sources to support gene-environment interaction studies.

This paper identifies potential vaccine candidates and drug targets for bovine tuberculosis using bioinformatics approaches and proteomic analysis of Mycobacterium bovis strains.

This study evaluates three docking programs for RNA-ligand screening, finding similar overall performance but differences across RNA classes and chemotypes, suggesting tailored scoring strategies could improve virtual screening.

This paper introduces RAMS, a framework that uses perceptual consistency to predict olfactory receptor activation spectra, improving accuracy and revealing new insights into OR function.

This paper introduces HaploDetox, a method for de novo bacterial haplotype reconstruction using a Conditional Random Field approach on de Bruijn graphs, achieving improved SNP phasing accuracy.

CausalKnowledgeTrace constructs causal graphs and identifies confounders and bias structures from biomedical literature using structured knowledge, improving variable selection for causal inference.