综合生物信息学、机器学习和实验验证确定了与PI3K/AKT信号相关的宫颈癌症的四基因诊断特征
Integrated bioinformatics, machine learning, and experimental validation identify a four-gene diagnostic signature for cervical cancer associated with PI3K/AKT signaling.
摘要(中文译文)
宫颈癌的早期准确诊断仍然是临床管理中的一大挑战。本研究旨在通过整合生物信息学和机器学习方法，识别宫颈癌可靠的诊断生物标志物，并进一步通过实验验证其生物学相关性。研究分析了来自基因表达综合数据库（Gene Expression Omnibus）和癌症基因组图谱（The Cancer Genome Atlas）的转录组数据，采用差异表达分析、加权基因共表达网络分析以及三种机器学习算法识别核心基因。通过受试者工作特征曲线和列线图模型评估诊断性能。通过药物敏感性分析、ssGSEA、免疫浸润分析和单细胞RNA测序探讨功能相关性。在10对宫颈癌及癌旁正常组织中进行了RT-qPCR验证，在3对组织样本中进行了Western blotting验证。使用SiHa和HeLa细胞进行体外验证。研究识别出CCND1、TRIP13、MYBL2和GNB4四个基因作为潜在的诊断生物标志物，联合模型显示出优于任何单一基因的诊断性能（AUC = 0.989）。3-甲基腺嘌呤处理改变了这些基因的表达，提示其可能与PI3K/AKT相关通路活性存在关联。此外，siRNA介导的GNB4敲低抑制了宫颈癌细胞增殖并降低了PI3K和AKT的磷酸化水平，为GNB4在PI3K/AKT通路激活中的功能参与提供了初步证据。CCND1、TRIP13、MYBL2和GNB4可能作为宫颈癌有前景的诊断生物标志物。其表达失调与PI3K/AKT通路活性相关，可能反映了参与宫颈癌进展的分子改变。特别是GNB4显示出潜在的诊断相关性和初步的功能意义，提示其可能作为候选生物标志物和分子靶点值得进一步研究。
Abstract
Early and accurate diagnosis remains a major challenge in cervical cancer management. This study aimed to identify reliable diagnostic biomarkers for cervical cancer by integrating bioinformatics and machine learning approaches and to further validate their biological relevance experimentally. Transcriptomic data from the Gene Expression Omnibus and The Cancer Genome Atlas were analyzed using differential expression analysis, weighted gene co-expression network analysis, and three machine learning algorithms to identify core genes. Diagnostic performance was evaluated using receiver operating characteristic curves and a nomogram model. Functional relevance was explored by drug sensitivity analysis, ssGSEA, immune infiltration analysis, and single-cell RNA sequencing. RT-qPCR validation was performed in 10 paired cervical cancer and adjacent normal tissues, while Western blotting was performed in three paired tissue samples. In vitro validation was conducted using SiHa and HeLa cells. Four genes, CCND1, TRIP13, MYBL2, and GNB4, were identified as potential diagnostic biomarkers, and the combined model showed superior diagnostic performance compared with any single gene (AUC = 0.989). Treatment with 3-methyladenine altered the expression of these genes, suggesting their potential association with PI3K/AKT-related pathway activity. Moreover, siRNA-mediated GNB4 knockdown suppressed cervical cancer cell proliferation and reduced PI3K and AKT phosphorylation, providing preliminary evidence for the functional involvement of GNB4 in PI3K/AKT pathway activation. CCND1, TRIP13, MYBL2, and GNB4 may serve as promising diagnostic biomarkers for cervical cancer. Their dysregulation was associated with PI3K/AKT pathway activity and may reflect molecular alterations involved in cervical cancer progression. In particular, GNB4 showed potential diagnostic relevance and preliminary functional significance, suggesting that it may represent a candidate biomarker and molecular target for further investigation.
如何引用
AMA
Hailong Zhang, Luhong Xie, Yameng Liu, Hanlin Yang, Shaoju Min, Yurong Zhu, et al. Integrated bioinformatics, machine learning, and experimental validation identify a four-gene diagnostic signature for cervical cancer associated with PI3K/AKT signaling.. Scientific reports. 2026; doi:10.1038/s41598-026-57689-7.
APA
Hailong Zhang, Luhong Xie, Yameng Liu, Hanlin Yang, Shaoju Min, Yurong Zhu, et al (2026). Integrated bioinformatics, machine learning, and experimental validation identify a four-gene diagnostic signature for cervical cancer associated with PI3K/AKT signaling.. Scientific reports. https://doi.org/10.1038/s41598-026-57689-7
GB/T 7714
Hailong Zhang, Luhong Xie, Yameng Liu, Hanlin Yang, Shaoju Min, Yurong Zhu, et al. Integrated bioinformatics, machine learning, and experimental validation identify a four-gene diagnostic signature for cervical cancer associated with PI3K/AKT signaling.[J]. Scientific reports, 2026 doi:10.1038/s41598-026-57689-7.
在 PubMed增强版 阅读全文
翻译全文 · AI 文献问答 · PDF 原文下载 · 引用导出