Identification and Validation of Biomarkers in Metabolic Dysfunction-Associated Steatohepatitis Using Machine Learning and Bioinformatics.

IF 1.5 4区医学 Q4 GENETICS & HEREDITY

Molecular Genetics & Genomic Medicine Pub Date : 2025-02-01 DOI:10.1002/mgg3.70063

Yu-Ying Zhang, Jin-E Li, Hai-Xia Zeng, Shuang Liu, Yun-Fei Luo, Peng Yu, Jian-Ping Liu

{"title":"Identification and Validation of Biomarkers in Metabolic Dysfunction-Associated Steatohepatitis Using Machine Learning and Bioinformatics.","authors":"Yu-Ying Zhang, Jin-E Li, Hai-Xia Zeng, Shuang Liu, Yun-Fei Luo, Peng Yu, Jian-Ping Liu","doi":"10.1002/mgg3.70063","DOIUrl":null,"url":null,"abstract":"Background: The incidence of metabolic dysfunction-associated steatohepatitis (MASH) is increasing annually. MASH can progress to cirrhosis and hepatocellular carcinoma. However, the early diagnosis of MASH is challenging.Aim: To screen prospective biomarkers for MASH and verify their effectiveness through in vitro and in vivo experiments.Methods: Microarray datasets (GSE89632, GSE48452, and GSE63067) from the Gene Expression Omnibus database were used to identify differentially expressed genes (DEGs) between patients with MASH and healthy controls. Machine learning methods such as support vector machine recursive feature elimination and least absolute shrinkage and selection operator were utilized to identify optimum feature genes (OFGs). OFGs were validated using the GSE66676 dataset. CIBERSORT was utilized to illustrate the variations in immune cell abundance between patients with MASH and healthy controls. The correlation between OFGs and immune cell populations was evaluated. The OFGs were validated at both transcriptional and protein levels.Results: Initially, 37 DEGs were identified in patients with MASH compared with healthy controls. In the enrichment analysis, the DEGs were mainly related to inflammatory responses and immune signal-related pathways. Subsequently, using machine learning algorithms, five genes (FMO1, PEG10, TP53I3, ME1, and TRHDE) were identified as OFGs. The candidate biomarkers were validated in the testing dataset and through experiments with animal and cell models. The malic enzyme (ME1) gene (HGNC:6983) expression was significantly upregulated in MASH samples compared to controls (0.4353 ± 0.2262 vs. -0.06968 ± 0.3222, p = 0.00076). Immune infiltration analysis revealed a negative correlation between ME1 expression and plasma cells (R = -0.77, p = 0.0033).Conclusion: This study found that ME1 plays a regulatory role in early MASH, which may affect disease progression by mediating plasma cells and T cells gamma delta to regulate immune microenvironment. This finding provides a new idea for the early diagnosis, monitoring and potential therapeutic intervention of MASH.","PeriodicalId":18852,"journal":{"name":"Molecular Genetics & Genomic Medicine","volume":"13 2","pages":"e70063"},"PeriodicalIF":1.5000,"publicationDate":"2025-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11850758/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Molecular Genetics & Genomic Medicine","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1002/mgg3.70063","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"GENETICS & HEREDITY","Score":null,"Total":0}

引用次数: 0

Abstract

Background: The incidence of metabolic dysfunction-associated steatohepatitis (MASH) is increasing annually. MASH can progress to cirrhosis and hepatocellular carcinoma. However, the early diagnosis of MASH is challenging.

Aim: To screen prospective biomarkers for MASH and verify their effectiveness through in vitro and in vivo experiments.

Methods: Microarray datasets (GSE89632, GSE48452, and GSE63067) from the Gene Expression Omnibus database were used to identify differentially expressed genes (DEGs) between patients with MASH and healthy controls. Machine learning methods such as support vector machine recursive feature elimination and least absolute shrinkage and selection operator were utilized to identify optimum feature genes (OFGs). OFGs were validated using the GSE66676 dataset. CIBERSORT was utilized to illustrate the variations in immune cell abundance between patients with MASH and healthy controls. The correlation between OFGs and immune cell populations was evaluated. The OFGs were validated at both transcriptional and protein levels.

Results: Initially, 37 DEGs were identified in patients with MASH compared with healthy controls. In the enrichment analysis, the DEGs were mainly related to inflammatory responses and immune signal-related pathways. Subsequently, using machine learning algorithms, five genes (FMO1, PEG10, TP53I3, ME1, and TRHDE) were identified as OFGs. The candidate biomarkers were validated in the testing dataset and through experiments with animal and cell models. The malic enzyme (ME1) gene (HGNC:6983) expression was significantly upregulated in MASH samples compared to controls (0.4353 ± 0.2262 vs. -0.06968 ± 0.3222, p = 0.00076). Immune infiltration analysis revealed a negative correlation between ME1 expression and plasma cells (R = -0.77, p = 0.0033).

Conclusion: This study found that ME1 plays a regulatory role in early MASH, which may affect disease progression by mediating plasma cells and T cells gamma delta to regulate immune microenvironment. This finding provides a new idea for the early diagnosis, monitoring and potential therapeutic intervention of MASH.

查看原文本刊更多论文

利用机器学习和生物信息学鉴定和验证代谢功能障碍相关脂肪性肝炎的生物标志物。

背景：代谢功能障碍相关脂肪性肝炎（MASH）的发病率逐年上升。MASH可发展为肝硬化和肝细胞癌。然而，早期诊断MASH是具有挑战性的。目的：筛选MASH的前瞻性生物标志物，并通过体内外实验验证其有效性。方法：采用基因表达综合数据库（Gene Expression Omnibus database）中的微阵列数据集（GSE89632、GSE48452和GSE63067），鉴定MASH患者与健康对照之间的差异表达基因（DEGs）。利用支持向量机递归特征消除、最小绝对收缩和选择算子等机器学习方法识别最优特征基因。ofg使用GSE66676数据集进行验证。使用CIBERSORT来说明免疫细胞丰度在MASH患者和健康对照之间的变化。评估了OFGs与免疫细胞群之间的相关性。在转录和蛋白水平上验证了ofg。结果：最初，与健康对照者相比，在MASH患者中鉴定出37个deg。在富集分析中，deg主要与炎症反应和免疫信号相关通路有关。随后，利用机器学习算法，鉴定出5个基因（FMO1、PEG10、TP53I3、ME1和TRHDE）为OFGs。候选生物标志物在测试数据集和动物和细胞模型实验中得到验证。与对照组相比，MASH样品中苹果酸酶（ME1）基因（HGNC:6983）表达显著上调（0.4353±0.2262 vs -0.06968±0.3222,p = 0.00076）。免疫浸润分析显示ME1表达与浆细胞呈负相关（R = -0.77, p = 0.0033）。结论：本研究发现ME1在早期MASH中发挥调控作用，可能通过介导浆细胞和T细胞γ δ调节免疫微环境影响疾病进展。这一发现为MASH的早期诊断、监测和潜在的治疗干预提供了新的思路。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Molecular Genetics & Genomic Medicine Biochemistry, Genetics and Molecular Biology-Genetics

CiteScore

4.20

自引率

0.00%

发文量

241

审稿时长

14 weeks

期刊介绍： Molecular Genetics & Genomic Medicine is a peer-reviewed journal for rapid dissemination of quality research related to the dynamically developing areas of human, molecular and medical genetics. The journal publishes original research articles covering findings in phenotypic, molecular, biological, and genomic aspects of genomic variation, inherited disorders and birth defects. The broad publishing spectrum of Molecular Genetics & Genomic Medicine includes rare and common disorders from diagnosis to treatment. Examples of appropriate articles include reports of novel disease genes, functional studies of genetic variants, in-depth genotype-phenotype studies, genomic analysis of inherited disorders, molecular diagnostic methods, medical bioinformatics, ethical, legal, and social implications (ELSI), and approaches to clinical diagnosis. Molecular Genetics & Genomic Medicine provides a scientific home for next generation sequencing studies of rare and common disorders, which will make research in this fascinating area easily and rapidly accessible to the scientific community. This will serve as the basis for translating next generation sequencing studies into individualized diagnostics and therapeutics, for day-to-day medical care. Molecular Genetics & Genomic Medicine publishes original research articles, reviews, and research methods papers, along with invited editorials and commentaries. Original research papers must report well-conducted research with conclusions supported by the data presented.