Identification and validation of oxidative stress-related immune signatures in Hepatitis B Virus associated acute liver failure based on WGCNA and machine learning
- 1Department of Infectious Diseases, Peking University Shenzhen hospital, Shenzhen, China
- 2Clinical Institute of Anhui Medical University & Peking University Shenzhen hospital, Shenzhen, China
- 3Department of Radiation Oncology, Guangzhou Institute of Cancer Research, the Affiliated Cancer Hospital, Guangzhou Medical University, Guangzhou, China
Abstract
Background: Acute liver failure (ALF) is a serious clinical disease. Immune infiltration and oxidative stress (OS) play an important role in ALF, but the combination of oxidative stress and immune infiltration in ALF has not been explored. Methods: Differential expression analysis and weighted gene co-expression network analysis (WGCNA) were performed on ALF and control samples in the GSE96851 dataset to obtain DE-OS-MGs. Candidate OS-IM-associated signature genes were mined using the least absolute shrinkage and selection operator (LASSO) and support vector machine (SVM) machine learning. Finally, OS-IM-related signature genes were obtained by expression analysis and ROC validation. Single gene set enrichment analysis (GSEA), TF-miRNA-mRNA network construction and drug prediction were also performed for OS-IM-related signature genes. Results: Differential analysis identified a total of 3580 DEGs in ALF and control samples, and differential analysis of immune cells yielded 15 immune cell species that were significantly different between the two groups. After WGCNA analysis, 1134 key module genes were screened. Then, 3580 DEGs, 855 OS-related genes, and 1134 key module genes were intersected, resulting in a total of 89 DE-OS-MGs. Furthermore, machine learning was performed and the genes were intersected to obtain 5 candidate OS-IM-related signature genes. The expression levels and ROC analysis of the 5 genes were validated in the GSE96851 and GSE120652 datasets, and 4 genes (CLU, APOH, AMBP, and GPX8) were used as biomarkers (OS-IM-related signature genes) for subsequent analysis. Single-gene GSEA enrichment analysis revealed that the 4 biomarkers were associated with metabolic processes. A miRNA-mRNA network with 123 nodes and 194 relationship pairs was obtained. 49 TFs of 3 biomarkers (APOH, AMBP, and GPX8) and 56 relationship pairs of TFs-mRNA regulatory network also constructed. Searching for therapeutic drugs of 4 biomarkers, a drug-mRNA network of CLU and 3 drugs was constructed. Conclusion: In this study, we combined immune infiltration landscape and oxidative stress based on bioinformatic analysis to screen for diagnostic markers (CLU, APOH, AMBP, and GPX8) of ALF. It provides a new idea and theoretical basis for the therapeutic diagnosis of ALF.
Related
- Identification of Crohn’s disease-associated molecular signatures by integrative bioinformatics analysis and experimental validation
- Single-cell and bulk transcriptome-based identification of myeloid-derived suppressor cell-associated immune evasion signatures and a prognostic model for gastric cancer.
- Integrative multi-omics, machine learning, computational toxicology, and experimental validation reveal PFAS-associated immune–inflammatory signatures in chronic obstructive pulmonary disease
- Identification and validation of biomarkers associated with mitochondrial dynamics dysfunction in Parkinson’s disease based on GEO datasets
- Identification and validation of cholesterol homeostasis and NK cell-associated predictive and therapeutic models for Coronary atherosclerotic heart disease patients with ST-elevated myocardial infarction: Insights from artificial intelligence and multi-omics
BibTeX
@article{dai2026identification,
title = {{Identification and validation of oxidative stress-related immune signatures in Hepatitis B Virus associated acute liver failure based on WGCNA and machine learning}},
author = {Lu Dai and Jungang Tan and Zhimin Zhou and Chengcong Chen and Guoxin Hu and Ling Guo},
journal = {PLOS One},
year = {2026},
month = sep,
doi = {10.1371/journal.pone.0356638},
url = {https://doi.org/10.1371/journal.pone.0356638},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Identification and validation of oxidative stress-related immune signatures in Hepatitis B Virus associated acute liver failure based on WGCNA and machine learning
AU - Lu Dai
AU - Jungang Tan
AU - Zhimin Zhou
AU - Chengcong Chen
AU - Guoxin Hu
AU - Ling Guo
PY - 2026
DA - 2026/09/09
JO - PLOS One
DO - 10.1371/journal.pone.0356638
UR - https://doi.org/10.1371/journal.pone.0356638
AB - Background: Acute liver failure (ALF) is a serious clinical disease. Immune infiltration and oxidative stress (OS) play an important role in ALF, but the combination of oxidative stress and immune infiltration in ALF has not been explored. Methods: Differential expression analysis and weighted gene co-expression network analysis (WGCNA) were performed on ALF and control samples in the GSE96851 dataset to obtain DE-OS-MGs. Candidate OS-IM-associated signature genes were mined using the least absolute shrinkage and selection operator (LASSO) and support vector machine (SVM) machine learning. Finally, OS-IM-related signature genes were obtained by expression analysis and ROC validation. Single gene set enrichment analysis (GSEA), TF-miRNA-mRNA network construction and drug prediction were also performed for OS-IM-related signature genes. Results: Differential analysis identified a total of 3580 DEGs in ALF and control samples, and differential analysis of immune cells yielded 15 immune cell species that were significantly different between the two groups. After WGCNA analysis, 1134 key module genes were screened. Then, 3580 DEGs, 855 OS-related genes, and 1134 key module genes were intersected, resulting in a total of 89 DE-OS-MGs. Furthermore, machine learning was performed and the genes were intersected to obtain 5 candidate OS-IM-related signature genes. The expression levels and ROC analysis of the 5 genes were validated in the GSE96851 and GSE120652 datasets, and 4 genes (CLU, APOH, AMBP, and GPX8) were used as biomarkers (OS-IM-related signature genes) for subsequent analysis. Single-gene GSEA enrichment analysis revealed that the 4 biomarkers were associated with metabolic processes. A miRNA-mRNA network with 123 nodes and 194 relationship pairs was obtained. 49 TFs of 3 biomarkers (APOH, AMBP, and GPX8) and 56 relationship pairs of TFs-mRNA regulatory network also constructed. Searching for therapeutic drugs of 4 biomarkers, a drug-mRNA network of CLU and 3 drugs was constructed. Conclusion: In this study, we combined immune infiltration landscape and oxidative stress based on bioinformatic analysis to screen for diagnostic markers (CLU, APOH, AMBP, and GPX8) of ALF. It provides a new idea and theoretical basis for the therapeutic diagnosis of ALF.
SN - 1932-6203
ER -