Yi Chuan. 2026 Sep;48(9):931-945. doi: 10.16288/j.yczz.25-275.

ABSTRACT

The abundance of omics data has significantly advanced the development of multi-omics data integration techniques. Non-linear embedding approaches for data integration have gradually become the mainstream in multi-omics research, as these approaches can substantially improve cancer analysis by enhancing the quality of the embeddings. However, current multi-omics data integration methods are typically confined to omics measurements, neglecting domain-specific prior knowledge encompassing biological pathways. In this study, we proposed a multi-omics integrated classification model, PathTransGCN, based on pathway self-attention and graph convolutional networks (GCN). The model integrated biological pathway information into multi-omics data analysis with the aim of enhancing the accuracy of cancer classification. Multi-omics data for breast cancer (BRCA), non-small cell lung cancer (NSCLC), and low-grade glioma (LGG) were obtained from The Cancer Genome Atlas (TCGA) and UCSC Xena databases. These data included gene mutations, DNA methylation, copy number variations, and gene expression, and were used to assess the model's generalizability across different cancers. First, PathTransGCN employed a pathway self-attention module to learn latent representations of samples across different pathways, thereby obtaining multi-omics integration vectors. Concurrently, a patient similarity network (PSN) was constructed using the similarity network fusion (SNF) approach. Second, the integrated vectors and the PSN were jointly fed into a GCN for end-to-end training, enabling precise classification of cancer subtypes. Through multi-omics data analysis of the BRCA dataset, PathTransGCN outperformed several popular algorithms (such as MoGCN and DeePathNet) in the five-class classification of cancer subtypes, achieving an accuracy rate of 87.6% and an F1 score of 86.4%. Moreover, the model demonstrated robust generalization capabilities across both NSCLC and LGG datasets, while effectively identifying key disease-associated biomarkers at the pathway level. Experimental results demonstrate that PathTransGCN exhibits outstanding performance in integrating omics data and delivering interpretable classification outcomes, presenting significant potential for clinical applications.

PMID:42751828 | DOI:10.16288/j.yczz.25-275

Front Immunol. 2026 Sep 2;17:1902053. doi: 10.3389/fimmu.2026.1902053. eCollection 2026.

ABSTRACT

Childhood asthma is a highly heterogeneous chronic respiratory disease, and its onset and progression are intricately linked to genetic susceptibility, environmental exposure, immune development, and the establishment of the early-life microbiome. In recent years, studies on the gut and respiratory microbiomes have suggested that the composition, metabolic functions, and interactions of microbial communities with the host immune system may be involved in the formation of asthma susceptibility, shaping of inflammatory phenotypes, and disease progression in children. The gut-lung axis, as an important pathway connecting gut microbiome, respiratory immunity, and systemic inflammatory responses, provides a new perspective for understanding the early mechanisms of childhood asthma. This article reviews the characteristics of the respiratory and gut microbiomes associated with childhood asthma, with a focus on the roles of the gut-lung axis, microbial metabolites, mucosal immune regulation, and environmental exposure. It also evaluates the research progress of probiotics, prebiotics, nutritional interventions, and novel microecological therapies. Additionally, the potential of microbial maturity, microbial metabolites, and immunophenotypes as biomarkers for risk prediction, phenotype stratification, and treatment response is analyzed. Furthermore, the role of multi-omics integration in supporting the identification of responsive populations, matching of intervention strategies, and dynamic monitoring of efficacy is discussed. Current evidence suggests that the microbiome offers promising targets for risk assessment and precision prevention of childhood asthma. However, relevant research still faces challenges such as ambiguous causality, high cohort heterogeneity, limited reproducibility of candidate biomarkers, inconsistent intervention outcomes, and insufficient evidence of long-term safety. At present, most biomarkers and multi-omics models remain in the stage of association discovery, lacking unified thresholds, cross-cohort validation, and biomarker-guided randomized controlled trials in children. Therefore, they cannot be routinely used for patient stratification or intervention selection. Future efforts should rely on standardized longitudinal birth cohorts, multi-omics integration, external validation, and high-quality clinical trials to clarify the incremental value of microbiome biomarkers over traditional clinical indicators and their clinical utility in the individualized management of childhood asthma.

PMID:42751182 | PMC:PMC13580037 | DOI:10.3389/fimmu.2026.1902053

Phenomics. 2026 May 20;6(3):213-229. doi: 10.1007/s43657-026-00322-4. eCollection 2026 Jun.

ABSTRACT

Early cancer detection remains challenging due to the lack of reliable pan-cancer screening methods, particularly blood-based biomarkers. Using a novel three-tiered validation framework combining artificial intelligence (AI)-powered literature mining of 180,000 PubMed articles (1950-2024), multiomics integration across major databases, and extensive clinical validation, we identified macrophage migration inhibitory factor (MIF) as a promising blood-based biomarker for pan-cancer detection. Multiomics analysis revealed consistent MIF upregulation across 21 cancer types at the transcriptional level and across 12 cancer types at the protein level. Clinical validation in independent cohorts (n = 4,269) showed that serum MIF protein levels discriminated effectively between cancer patients and healthy controls (median AUC = 0.994) and between cancer and benign conditions (median AUC = 0.881). Notably, comparative analyses showed that MIF demonstrated superior or comparable performance to established cancer-specific markers, including AFP for hepatocellular carcinoma (MIF AUC = 0.885 vs. AFP AUC: 0.744-0.887) and CA125 for ovarian cancer (MIF AUC = 0.831 vs. CA125 AUC: 0.58-0.71). Meta-analysis of 28 cohorts (n = 5,347) confirmed the diagnostic efficacy of MIF (pooled AUC: 0.782). This cost-effective, blood-based ELISA approach establishes MIF as a valuable tool for broad applications in cancer screening.

SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at https://doi.org/10.1007/s43657-026-00322-4.

PMID:42750739 | PMC:PMC13578188 | DOI:10.1007/s43657-026-00322-4

Mater Today Bio. 2026 Aug 18;40:103583. doi: 10.1016/j.mtbio.2026.103583. eCollection 2026 Oct.

ABSTRACT

TLR4 activation and the cascade of neutrophil extracellular trap (NET) formation exacerbate excessive inflammation and organ damage in the pathogenesis of acute lung injury (ALI), yet effective pharmacological interventions remain unavailable. Nanoaggregates derived from natural products offer promising avenue by leveraging synergistic anti-inflammatory effects. In this study, we surprisingly discovered that leonurine and ursolic acid spontaneously self-assemble into nanoparticles (LUNP) through non-covalent interactions, achieving a drug loading capacity of 100%. The LUNP platform exhibits superior biophysical properties, including enhanced mucus penetration, pH-responsive drug release, improved cellular uptake, and prolonged retention within inflamed lung tissue. Mechanistically, LUNP ameliorates ALI by dampening TLR4/MyD88/NF-κB-driven inflammatory activation, thereby remodeling the microenvironment to limit NOX4-PAD4-mediated NET formation. Notably, inhalational LUNP exhibits outstanding biosafety with minimal off-target distribution. Overall, this work introduces a synergistic self-assembled nanoplatform for precise pulmonary intervention in ALI, showcasing its ability to safely and effectively orchestrate the coordinated modulation of multiple pathological pathways. In summary, by inhibiting both TLR4 activation and NET formation, the synergistic LUNP platform offers an efficient, safe, and easily accessible therapeutic strategy for ALI, providing a promising solution for clinical translation.

PMID:42750707 | PMC:PMC13577835 | DOI:10.1016/j.mtbio.2026.103583

J Thorac Oncol. 2026 Sep 16:104204. doi: 10.1016/j.jtho.2026.104204. Online ahead of print.

ABSTRACT

BACKGROUND: Early-stage EGFR-mutant lung adenocarcinoma (LUAD) demonstrates heterogeneous outcomes after curative surgery, yet adjuvant treatment decisions are guided by pathological stage alone. Following the ADAURA trial, adjuvant osimertinib is the standard of care for resected stage IB-IIIA EGFR-mutant LUAD; however, real-world data demonstrate that up to 40% of patients remain disease-free at five years without adjuvant osimertinib, underscoring the need for improved risk stratification.

PATIENTS AND METHODS: We performed integrated clinical, genomic and transcriptomic profiling of 400 patients with resected stage IA-IIIA EGFR-mutant LUAD. EGFR-mutant recurrence risk models integrating clinical, genomic and transcriptomic data were developed and validated across one internal and three external cohorts.

RESULTS: Genomic instability, including TP53 co-mutations, copy number alterations and APOBEC-associated mutational signatures, increased with pathological stage. RBM10 co-mutations were enriched in tumours with L858R mutations and correlated with upregulation of WNT signalling and epithelial-mesenchymal transition. Transcriptomic features outperformed clinical or genomic variables alone in predicting recurrence risk, and a multi-omic model demonstrated superior and reproducible performance, achieving a median concordance index of 75.4% across four independent validation cohorts. The multi-omic model stratified recurrence risk within individual pathological stages, including stage I disease, and identified patients most likely to benefit from adjuvant EGFR TKI.

CONCLUSIONS: These findings define the molecular heterogeneity of early-stage EGFR-mutant LUAD and support multi-omic risk stratification to inform adjuvant EGFR TKI decisions beyond pathological stage. Prospective validation in larger cohorts will be required to confirm these findings.

PMID:42749051 | DOI:10.1016/j.jtho.2026.104204

Nat Genet. 2026 Sep 15. doi: 10.1038/s41588-026-02751-3. Online ahead of print.

ABSTRACT

Inflammation accelerates evolutionary dynamics of hematopoietic stem cells (HSCs) in clonal hematopoiesis and myeloid neoplasms. We studied HSCs, progenitors and immune cells from patients with myeloproliferative neoplasms at baseline and following interferon-α (IFNα) treatment, the only therapy to deplete mutated stem cells. We deployed single-cell multiomics methods that distinguish the IFNα effects on mutated stem cells from the admixed wild-type HSCs, with respect to their differentiation, transcriptomes, immunophenotypes and chromatin accessibility. IFNα simultaneously activated HSCs into two polarized states: a lymphoid progenitor expansion associated with an anti-inflammatory state and an inflammatory myeloid progenitor state derived from HSCs. The augmented lymphoid differentiation balanced the typical myeloproliferative-neoplasm-induced myeloid bias, associated with normalized blood counts. Somatic mutations modified the effects of IFNα on HSC differentiation and cell cycle entry rates. Clonal fitness upon IFNα exposure was due to resistance of CALR- or JAK2-mutated stem cells to differentiate into inflammatory myeloid progenitors.

PMID:42745000 | DOI:10.1038/s41588-026-02751-3