自发性早产中关键代谢相关基因和途径的鉴定：结合生物信息学分析和机器学习。

Identification of key metabolism-related genes and pathways in spontaneous preterm birth: combining bioinformatic analysis and machine learning.

机构信息

Department of Obstetrics, Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal-Fetal Medicine and Gynecologic Oncology, Shanghai, China.

Department of Clinical Medicine, Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal-Fetal Medicine and Gynecologic Oncology, Shanghai, sChina.

出版信息

Front Endocrinol (Lausanne). 2024 Aug 20;15:1440436. doi: 10.3389/fendo.2024.1440436. eCollection 2024.

DOI:10.3389/fendo.2024.1440436

PMID:39229380

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC11368757/

Abstract

BACKGROUND

Spontaneous preterm birth (sPTB) is a global disease that is a leading cause of death in neonates and children younger than 5 years of age. However, the etiology of sPTB remains poorly understood. Recent evidence has shown a strong association between metabolic disorders and sPTB. To determine the metabolic alterations in sPTB patients, we used various bioinformatics methods to analyze the abnormal changes in metabolic pathways in the preterm placenta via existing datasets.

METHODS

In this study, we integrated two datasets (GSE203507 and GSE174415) from the NCBI GEO database for the following analysis. We utilized the "Deseq2" R package and WGCNA for differentially expressed genes (DEGs) analysis; the identified DEGs were subsequently compared with metabolism-related genes. To identify the altered metabolism-related pathways and hub genes in sPTB patients, we performed multiple functional enrichment analysis and applied three machine learning algorithms, LASSO, SVM-RFE, and RF, with the hub genes that were verified by immunohistochemistry. Additionally, we conducted single-sample gene set enrichment analysis to assess immune infiltration in the placenta.

RESULTS

We identified 228 sPTB-related DEGs that were enriched in pathways such as arachidonic acid and glutathione metabolism. A total of 3 metabolism-related hub genes, namely, ANPEP, CKMT1B, and PLA2G4A, were identified and validated in external datasets and experiments. A nomogram model was developed and evaluated with 3 hub genes; the model could reliably distinguish sPTB patients and term labor patients with an area under the curve (AUC) > 0.75 for both the training and validation sets. Immune infiltration analysis revealed immune dysregulation in sPTB patients.

CONCLUSION

Three potential hub genes that influence the occurrence of sPTB through shadow participation in placental metabolism were identified; these results provide a new perspective for the development and targeting of treatments for sPTB.

摘要

背景

自发性早产（sPTB）是一种全球性疾病，是导致新生儿和 5 岁以下儿童死亡的主要原因。然而，sPTB 的病因仍知之甚少。最近的证据表明代谢紊乱与 sPTB 之间存在很强的关联。为了确定 sPTB 患者的代谢变化，我们使用各种生物信息学方法通过现有的数据集分析早产胎盘代谢途径的异常变化。

方法

在这项研究中，我们整合了 NCBI GEO 数据库中的两个数据集（GSE203507 和 GSE174415）进行如下分析。我们利用“Deseq2”R 包和 WGCNA 进行差异表达基因（DEGs）分析；鉴定的 DEGs 随后与代谢相关基因进行比较。为了确定 sPTB 患者中改变的代谢相关途径和枢纽基因，我们进行了多种功能富集分析，并应用了三种机器学习算法，LASSO、SVM-RFE 和 RF，以及通过免疫组化验证的枢纽基因。此外，我们还进行了单样本基因集富集分析，以评估胎盘中的免疫浸润。

结果

我们确定了 228 个与 sPTB 相关的 DEGs，这些基因富集在花生四烯酸和谷胱甘肽代谢等途径中。总共确定了 3 个与代谢相关的枢纽基因，即 ANPEP、CKMT1B 和 PLA2G4A，并在外部数据集和实验中进行了验证。我们开发了一个基于 nomogram 模型，并使用 3 个枢纽基因进行评估；该模型在训练集和验证集的 AUC>0.75 时，可以可靠地区分 sPTB 患者和足月分娩患者。免疫浸润分析显示 sPTB 患者存在免疫失调。