This study developed an 11-gene lymph node metastasis gene signature (LMGS) using an ensemble learning strategy to overcome the instability of traditional models and identifies SPIB as a molecular marker associated with lymphatic dissemination.
Abstract
Summary The heterogeneity within the same AJCC TNM stage of gastric cancer necessitates robust prognostic biomarkers beyond conventional staging. By integrating transcriptomic data from 1,613 patients across eight cohorts, we developed an 11-gene lymph node metastasis gene signature (LMGS) using an ensemble learning strategy to overcome the instability of traditional models. LMGS robustly predicted survival (mean C-index = 0.619), providing incremental prognostic value beyond conventional AJCC staging. Single-cell and spatial transcriptomic analyses revealed LMGS enrichment in epithelial/immune cells at the tumor-normal interface. Integrative analysis nominated the transcription factor SPIB as a candidate regulator, which was specifically overexpressed in metastatic lesions. Our study provides a clinically relevant prognostic tool and identifies SPIB as a molecular marker associated with lymphatic dissemination.
This study establishes a novel CSC–associated gene signature for diagnosis and prognosis in HCC and nominates belinostat as a repurposing candidate for targeting stemness‐related pathways, offering a promising strategy for personalized therapy.
Yang Zi, Ying Zhang, Jun Wu et al.· Stem Cells International· 0 citations
A dual-gene model based on neutrophil heterogeneity demonstrated strong predictive performance and functioned as an independent prognostic indicator, and distinct immune and prognostic differences were identified among molecular subtypes.
Background Lymph node metastasis (LNM) is a key prognostic determinant in thyroid carcinoma (THCA), yet molecular markers capturing intrinsic metastatic potential are limited. Methods Generalized additive models were applied to TCGA-THCA data to screen for genes with diametrically opposite expression–tumor diameter relationships between N0 and N1 patients. CENPM was subsequently validated in independent transcriptomic cohorts, spatial transcriptomics, and immunohistochemistry. Single-cell transcriptomics, in silico knockout, drug repositioning, and molecular docking were employed to dissect its immunological roles and therapeutic relevance. Results CENPM expression increased with tumor diameter in N0 but decreased in N1, and high CENPM was associated with poorer disease-free survival. CENPM was predominantly enriched in CD8+ naïve and effector T cells, particularly in anaplastic carcinoma. Virtual knockout predicted downstream transcriptional changes associated with lymphocyte activation, translational machinery, and immune effector pathways. Drug repositioning identified filgotinib as a candidate to reverse the CENPM-high signature, with stable CENPM–filgotinib binding confirmed by docking and molecular dynamics. Conclusion CENPM shows diametrically opposite expression–diameter relationships between N0 and N1 patients, suggesting a shift in biological behavior upon nodal involvement, with potential therapeutic relevance.
SIRPG is identified as an immune-related prognostic hub and context-dependent tumor-cell regulator associated with apoptosis, immune communication and spatial microenvironmental organization in HNSCC.
Jiaqi Tang, Yu-Lun He, Yuqi Wang et al.· Frontiers in Immunology· 0 citations
Our study developed a multiomics-driven transformer model that combines mammography, MRI, transcriptomic and proteomic data to noninvasively predict axillary lymph node (ALN) metastasis in breast cancer. A total of 2105 patients from 10 institutions were included for model training and validation. The model achieved an AUC of 0.939 in the training cohort (n = 658) and 0.830-0.867 across three independent validation cohorts (n = 282, 971 and 194, respectively), outperforming conventional ultrasound examination. Grad-CAM visualizations highlighted the tumor edges and surrounding tissue, consistent with clinical and pathological findings. In a cohort of 194 patients, multiomics analyses linked the model output to gene and protein signatures involved in immune modulation, cytoskeletal remodeling, and epithelial-to-mesenchymal transition. Critically, the major enriched pathways identified through model-stratified analysis were independently replicated in a parallel non-model-driven analysis using ALN status, demonstrating that these signatures reflect tumor biology. Network analysis revealed gene clusters related to DNA replication and immune pathways, providing biological insights into the model's decisions. These findings suggest that the stacking model holds promise as a noninvasive decision-support tool that may complement, rather than replace, current clinical staging practices. However, integration into clinical workflows requires prospective validation.
Xiaodong Liu, Fan Li, Ye Xiang et al.· npj Precision Oncology· 0 citations
An interpretable three-stage machine learning framework integrating mRNA, microRNA, DNA methylation, copy number variation, and protein expression data from The Cancer Genome Atlas that couples improved prognostic estimation with biological interpretability supporting multi-omics biomarker discovery in ovarian cancer.
Kosar Fateh, S. Sathipati· bioRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
