Wang Dan, Huo Rujie, Ye Lu
Department of Respiratory Medicine, The Second Hospital of Shanxi Medical University, 382 Wuyi Road, Xinghualing Area, 030000, Taiyuan, China.
Heliyon. 2024 Dec 4;11(1):e40870. doi: 10.1016/j.heliyon.2024.e40870. eCollection 2025 Jan 15.
N7-methylguanosine (m7G) modification is closely related to the occurrence of human diseases, but its roles in sepsis remain unclear. This study aimed to explore the patterns of lethality-related m7G regulatory factor-mediated RNA methylation modification and immune microenvironment regulatory features in sepsis.
Three sepsis-related datasets (E-MTAB-4421 and E-MTAB-4451 as training sets and GSE185263 as a validation set) were collected, and differentially expressed m7G-related genes were analyzed between survivors and non-survivors. Lethality-related m7G signature genes were then screened using machine learning methods, followed by the construction of a survival recognition model. Additionally, differences in immune cell distribution were determined and differentially expressed genes (DEGs) between different subtypes were analyzed. Weighted gene co-expression network analysis (WGCNA) was used to select important modules and related hub genes.
In total, 10 differentially expressed m7G-related genes were identified between the survivors and non-survivors, and after further analysis, , , , , and were identified as the optimal lethality-related m7G genes. A survival status diagnostic model was then constructed with a combined AUC of 0.678. Fifteen types of immune cells were significantly different between survivors and non-survivors. Sepsis samples were classified into two subtypes, with 22 types of immune cells showing significant differences. Subsequently, 1707 DEGs were identified between the two subtypes, which were significantly enriched in 91 GO terms and 16 KEGG pathways. Finally, the green module with |correlation| > 0.3 was found to be closely related to the subtypes and survival status; further, the top10 hub genes were obtained.
The constructed survival status diagnostic model based on the five lethality-related m7G signature genes may help predict the survival status of patients, and the 10 hub genes obtained may be potential therapeutic targets for sepsis.
N7-甲基鸟苷（m7G）修饰与人类疾病的发生密切相关，但其在脓毒症中的作用仍不清楚。本研究旨在探讨脓毒症中与致死率相关的m7G调控因子介导的RNA甲基化修饰模式及免疫微环境调控特征。
收集三个脓毒症相关数据集（E-MTAB-4421和E-MTAB-4451作为训练集，GSE185263作为验证集），分析幸存者和非幸存者之间差异表达的m7G相关基因。然后使用机器学习方法筛选与致死率相关的m7G特征基因，随后构建生存识别模型。此外，确定免疫细胞分布差异并分析不同亚型之间的差异表达基因（DEG）。使用加权基因共表达网络分析（WGCNA）选择重要模块和相关枢纽基因。
共鉴定出幸存者和非幸存者之间10个差异表达的m7G相关基因，进一步分析后， 、 、 、 、 和 被确定为最佳的与致死率相关的m7G基因。然后构建了一个生存状态诊断模型，其组合AUC为0.678。幸存者和非幸存者之间有15种免疫细胞存在显著差异。脓毒症样本分为两个亚型，22种免疫细胞存在显著差异。随后，在两个亚型之间鉴定出1707个DEG，这些DEG在91个GO术语和16条KEGG通路中显著富集。最后，发现相关性|>0.3|的绿色模块与亚型和生存状态密切相关；此外，获得了前10个枢纽基因。
基于五个与致死率相关的m7G特征基因构建的生存状态诊断模型可能有助于预测患者的生存状态，获得的10个枢纽基因可能是脓毒症的潜在治疗靶点。
