实用医学杂志 ›› 2024, Vol. 40 ›› Issue (9): 1230-1237.doi: 10.3969/j.issn.1006-5725.2024.09.009

• 临床研究 • 上一篇    下一篇

m1A/m5C/m6A/m7G调控基因预测胃癌预后及免疫关联性

陈小梅1,2,王安奇1,2,杨积祯3,于淼1()   

  1. 1.甘肃省人民医院 国家胃肠肿瘤诊治重点实验室
    2.甘肃省人民医院 临床研究与转化医学研究所
    3.甘肃省人民医院 超声医学科 (兰州 730000 )
  • 收稿日期:2024-01-08 出版日期:2024-05-10 发布日期:2024-05-15
  • 通讯作者: 于淼 E-mail:travy@126.com
  • 基金资助:
    国家卫生健康委胃肠肿瘤诊断与治疗重点实验室2022级硕士/博士/博士后基金资助项目(NHCDP2022014);甘肃省自然科学基金资助项目(23JRRA1770);甘肃省人民医院院内基金青年培育资助项目(23GSSYF-16)

Prognosis and immune correlation analysis of m1A/m5C/m6A/m7G regulated genes in gastric cancer

Xiaomei CHEN1,2,Anqi WANG1,2,Jizhen YANG3,Miao YU1()   

  1. National Key Laboratory of Gastrointestinal Tumor Diagnosis and Treatmen,Institute of Clinical Research and Translational Medicine,Gansu Provincial Hospital,Lanzhou 730000,China
  • Received:2024-01-08 Online:2024-05-10 Published:2024-05-15
  • Contact: Miao YU E-mail:travy@126.com

摘要:

目的 基于m1A/m5C/m6A/m7G甲基化调控基因建立胃癌预后风险预测模型,并分析该模型与免疫的关联性。 方法 通过癌症基因组图谱(TCGA)-胃癌数据集筛选表达具有显著差异的m1A/m5C/m6A/m7G调控基因,通过单基因Cox回归分析和LASSO算法构建预后风险评分(risk score, RS)模型,使用Kaplan-Meier(K-M)统计进行RS模型验证,并应用细胞系进行RT-qPCR验证。利用单因素、多因素Cox回归分析建立nomogram模型。使用CIBERSORT算法和estimate包进行免疫关联性分析。 结果 建立了基于八个甲基化调控基因的预后RS模型,将胃癌患者分为高风险和低风险。这八个基因在胃癌细胞系中高表达(P < 0.05)。在TCGA-胃癌训练集和GSE62254-验证集中,患者总生存率 (OS)与分组状态之间存在显著相关性(P < 0.001)。nomogram生存模型预测的1年(C-index=0.703)、3年(C-index=0.729)和5年(C-index=0.734)生存率与实际生存率的一致性较好。免疫关联性分析表明,与低风险患者组相比,高风险患者组免疫评分和免疫检查点相关基因的表达较高(P < 0.05)。 结论 基于m1A/m5C/m6A/m7G调控基因的预后RS模型可预测胃癌预后并指导个体免疫治疗决策。

关键词: m1A, m5C, m6A, m7G, 胃癌, 预后, 免疫浸润

Abstract:

Objective This study aims to develop a prognostic risk prediction model for gastric cancer based on m1A/m5C/m6A/m7G regulated genes and to investigate the relationship between this model and immunology. Methods The Cancer Genome Atlas (TCGA) gastric cancer dataset was utilized to identify m1A/m5C/m6A/m7G regulated genes with significant expression differences. A prognostic risk score (RS) model was constructed using univariate Cox regression analysis and the LASSO algorithm. The RS model was validated using the Kaplan?Meier (K?M) statistic and cell lines for RT?qPCR biological validation. A nomogram model was created using univariate and multivariate Cox regression analyses. The CIBERSORT algorithm and ESTIMATE package were employed to conduct immune correlation analysis. Results A prognostic RS model based on eight methylation regulated genes was developed to classify patients with gastric cancer as high?risk or low?risk. These eight genes showed significant expression in gastric cancer cell lines (P < 0.05). The TCGA?gastric cancer training set and GSE62254?validation set showed a substantial connection (P < 0.001) between overall survival rate (OS) and grouping status. The nomogram survival models accurately predicted 1?year (C?index = 0.703), 3?year (C?index = 0.729), and 5?year (C?index = 0.734) survival rates. Immune correlation analysis showed that compared to the low?risk group, the high?risk group had higher immune scores and higher expression of immune checkpoint?related genes (P < 0.05). Conclusion We created a reliable prognostic RS model based on m1A/m5C/m6A/m7G regulated genes that can predict gastric cancer prognosis and guide individualized immunotherapy decisions.

Key words: m1A, m5C, m6A, m7G, gastric cancer, prognosis, immune infiltration

中图分类号: