Discover, rate, and discuss biology preprints.

The paper fits the Gompertz curve to PDB growth data to estimate future PDB entries and protein folds, revealing a deceleration in growth.

The paper develops deep-learning models to predict glycan-protein binding strengths using hybrid datasets and tail-aware losses, finding that glycan interactions are harder than small-molecule interactions due to chiral feature limitations.

CoLa-VAE improves denoising and representation learning in single-cell data by integrating cell-cell communication information, enhancing downstream analyses like differential communication detection and spatial deconvolution.

FoldARE improves RNA secondary structure prediction by using pseudo-SHAPE constraints from structural ensembles, achieving high accuracy across benchmark datasets.

This study analyzes a diverse glioma cohort to show ancestry influences on survival, somatic alterations, and mutational processes in diffuse gliomas.

BGC-QDR is a quantum-assisted pipeline for discovering and ranking biosynthetic gene clusters from eDNA, achieving 0.789 accuracy and 0.835 ROC-AUC using a variational quantum classifier.

CREPAS is a Nextflow pipeline for analyzing nascent chromatin sequencing data, improving scalability, reproducibility, and compatibility across multiple chromatin assays.

The paper introduces DSEN, a tissue-aware regression method that improves drug response prediction by decomposing models into shared and lineage-specific components, achieving better performance and interpretability.

The paper introduces an ambiguity-aware framework for spatial cell-type annotation, reducing cluster and cell-level ambiguity in cholangiocarcinoma data.

This paper introduces a computational pipeline using DNA k-mer imprints to agnostically classify and distinguish materials based on their sequence patterns.

CellOS is a multi-view foundation model trained on single-cell transcriptomes that outperforms existing models in cell-state annotation, batch integration, and perturbation-response prediction.

NanoCellAnnotator improves spatial transcriptomics cell-type annotation by combining biological constraints and LLMs, identifying canonical cell populations and flagging ambiguous clusters.

Bamsnap-LRS automates batch visualization of long-read sequencing alignments, supporting genomic, transcriptomic, and haplotype-aware analyses with publication-ready figure generation.

DextraDemixer improves identification of antigen-specific T cells from pMHC multimer data by using a Bayesian model and dynamic thresholding to reduce false discoveries.

The paper constructs a healthy human pancreas reference atlas to contextualize disease states and benchmark models.

ATAClone identifies cancer clones and estimates copy numbers from scATAC-seq data, improving accuracy over existing methods.

This study analyzes 76 CALR exon 9 variants to refine their classification based on charge architecture, revealing subgroups within Type-1-like variants and linking tail chemistry to receptor activation.

BATTLE-AMP benchmarks antimicrobial peptide predictors using MIC data, showing that models trained on MIC data outperform binary classifiers and highlighting limitations in distinguishing active from inactive peptides.

This paper introduces a framework to stratify pharmacological data by action type and assay readout, improving QSAR model validity for opioid receptors.

This paper introduces a method to expand gene regulatory networks by incorporating prior knowledge, outperforming conventional methods in distinguishing true interactions from indirect associations.