🧬 biology

Disentangling Loss Design and Ontology-Aware Architecture in GNN-Based Protein Function Prediction A Controlled Ablation Study

This controlled ablation study reveals that in GNN-based protein function prediction, the proposed Information Accretion-aware loss function is counterproductive, whereas the optimal performance is achieved by combining a GCN architecture with cross-ontology attention and standard binary cross-entropy loss.

Yongjin Chen, Xiyuan Wang, Yiman Gao, Songze Zhu2026-09-10
🧬 biology

H18 haemagglutinin binds MHC class II at a site distant from the canonical sialic-acid binding site

This study reveals that the bat-derived influenza H18 haemagglutinin binds to MHC class II molecules at a novel site distinct from the canonical sialic-acid pocket, inducing conformational changes that prime viral fusion and demonstrating the protein's structural adaptability to utilize protein receptors for host entry.

Antoni Wrobel, Jonathan Robert, Maria Osman, Valeria Calvaresi, Benjamin Bouzabia, Selene Franchini, Sophie Focks, Liubo (…)2026-09-10
🧬 biology

Evaluating porcine antisense oligonucleotide RNA-seq for cross- species virtual-knockout benchmarking

This study evaluates porcine antisense oligonucleotide RNA-seq data to determine the viability of cross-species virtual-knockout benchmarking, concluding that while whole-transcriptome scoring is unreliable due to estimator discrepancies, leading-gene rank scoring is reproducible and offers a defensible target for extending human-derived perturbation models to livestock.

Xiaotong Zhao, Hua Chang, Zhuoyu Zhao, Xun Xiang2026-09-10
🧬 biology

Geographic Variation in the Urinary Microbiota of Patients with Urolithiasis

This cross-national study reveals that while urinary microbial richness is similar, the overall community composition and specific taxonomic abundances in patients with urolithiasis differ significantly between Taiwan and South Korea, suggesting that geographic and environmental factors shape distinct urinary microbiota profiles associated with stone disease.

Hsiang Ying Lee, Chun Kai Chang, Wen-Jeng Wu, Deng‐Chyang Wu, Sung Yong Cho2026-09-10
🧬 biology

Adaptation and psychometric evaluation of the BANKS-SP biobanking survey in longstanding dengue virus cohorts in Colombia and Nicaragua

This study adapts and evaluates the psychometric properties of the BANKS-SP biobanking survey in Colombian and Nicaraguan dengue cohorts, finding that while the Self-efficacy Scale performed adequately, the Attitudes Scale requires revision and both sites exhibited significant gaps in knowledge regarding biospecimen governance and legal frameworks.

Lauren Maxwell, Mahir Bhatt, Harold Suazo Laguna, Jackeline Bravo Chamorro, María Consuelo Miranda Montoya, Luz Leegstra (…)2026-09-10
🧬 biology

Umbilical Cord Mesenchymal Stem Cells Alleviate Osteoblast Differentiation Disorder Induced by Oxidative Stress by Inhibiting Ferroptosis Through the miR-218-5p/CBX3 Axis

This study demonstrates that umbilical cord mesenchymal stem cells (UC-MSCs) alleviate oxidative stress-induced osteoblast differentiation disorders and inhibit ferroptosis by upregulating miR-218-5p, which targets and suppresses CBX3, thereby restoring osteogenic function via the miR-218-5p/CBX3 signaling axis.

Lintong Li, Yunda Huang, Yuan Ma, Jiaxin Liu, Yu Chang2026-09-10
🧬 biology

A novel topology-guided adaptive strategy for automated nuclear-cytoplasmic H-Score quantification in immunohistochemistry

This paper introduces an open-source, training-free ImageJ/Fiji plugin that combines StarDist nuclear detection with a topology-guided Voronoi-based segmentation strategy and adaptive calibration to achieve highly accurate, reproducible automated quantification of nuclear and cytoplasmic H-scores across diverse immunohistochemistry datasets.

Weiyu Wang, Mengting Lin, Fuxian Zou, Zhenzhen Yang, Mingqing Huang, Changxian Chen2026-09-10
🧬 biology

Reduced Representation Nanopore Sequencing: A Cost-Effective Workflow for Accurate and High-Throughput Targeted DNA Methylation Profiling

The study introduces Reduced Representation Nanopore Sequencing (RRNS), a cost-effective workflow that combines MspI-based targeted enrichment with direct nanopore sequencing to enable high-throughput, accurate DNA methylation profiling while overcoming the limitations of bisulfite conversion and high whole-genome sequencing costs.

Keisho Hasegawa, Daigo Okada, Kengo Suzuki, Marumi Ishikawa, Akinori Tsuruma, Masaki Kimura, Yoshimasa Saito2026-09-10