Al-Harazi Olfat, El Allali Achraf, Kaya Namik, Colak Dilek
Molecular Oncology Department, King Faisal Specialist Hospital and Research Centre, Riyadh 11211, Saudi Arabia.
Bioinformatics Laboratory, College of Computing, Mohammed VI Polytechnic University, Benguerir 43150, Morocco.
Int J Mol Sci. 2024 Nov 28;25(23):12779. doi: 10.3390/ijms252312779.
Breast cancer remains a major global health concern and a leading cause of cancer-related deaths among women. Early detection and effective treatment are essential in improving patient survival. Advances in omics technologies have provided deeper insights into the molecular mechanisms underlying breast cancer. This study aimed to identify subnetwork markers with diagnostic and prognostic potential by integrating genome-wide gene expression data with protein-protein interaction networks. We identified four significant subnetworks revealing potentially important hub genes, including , , , , , , , and . The diagnostic and prognostic potentials of these subnetworks were validated using independent datasets. Unsupervised principal component analysis demonstrated a clear separation of breast cancer patients from healthy controls across multiple datasets. A KNN classification model, based on these subnetworks, achieved an accuracy of 97%, sensitivity of 98%, specificity of 94%, and area under the curve (AUC) of 96%. Moreover, the prognostic significance of these subnetwork markers was validated using independent transcriptomic datasets comprising over 4000 patients. These findings suggest that subnetwork markers derived from integrated genomic network analyses can enhance our understanding of the molecular landscape of breast cancer, potentially leading to improved diagnostic, prognostic, and therapeutic strategies.
乳腺癌仍然是全球主要的健康问题，也是女性癌症相关死亡的主要原因。早期检测和有效治疗对于提高患者生存率至关重要。组学技术的进步为深入了解乳腺癌的分子机制提供了更多见解。本研究旨在通过整合全基因组基因表达数据与蛋白质-蛋白质相互作用网络，识别具有诊断和预后潜力的子网络标志物。我们确定了四个重要的子网络，揭示了潜在的重要枢纽基因，包括 ， ， ， ， ， ， 和 。这些子网络的诊断和预后潜力使用独立数据集进行了验证。无监督主成分分析表明，在多个数据集中，乳腺癌患者与健康对照有明显区分。基于这些子网络的KNN分类模型，准确率达到97%，灵敏度为98%，特异性为94%，曲线下面积（AUC）为96%。此外，使用包含4000多名患者的独立转录组数据集验证了这些子网络标志物的预后意义。这些发现表明，从整合基因组网络分析中得出的子网络标志物可以增强我们对乳腺癌分子格局的理解，有可能带来改进的诊断、预后和治疗策略。
