Identification and validation of mitochondrial-related genes in intestinal ischemia-reperfusion injury based on WGCNA and machine learning

YiChen Hu¹Jie Huang²XiaoLi Min²YuanPei Zhao¹JiaHui Wang³HongYuan Liu¹KaiWen Shi¹WenLiang Li^2*WeiMing Li^1*

• ¹Kunming Medical University Second Affiliated Hospital Department of Gastrointestinal Surgery Second Ward, Kunming, China
• ²The Second Affiliated Hospital of Kunming Medical University, Kunming, China
• ³Yunnan Cancer Hospital, Kunming, China

Background: Severe ischemia-reperfusion (II/R) injury of the intestines is a leading cause of death and disability. According to earlier research, modulating mitochondrial function is the primary mechanism by which II/R injury is ameliorated. In order to further molecular diagnostics and discover possible treatment targets, it is essential to find biomarkers of mitochondria in II/R injury. Methods: The datasets GSE96733 and GSE37013, along with mitochondrial-related genes (MRGs), were obtained from the Gene Expression Omnibus (GEO) database and MitoCarta3.0, respectively. GSE96733 conducted differential expression gene (DEGs) analysis and weighted gene co-expression network analysis (WGCNA) module screening. In order to find MRGs that were expressed differently, we got their intersection (DEMRGs) and gene enrichment analysis was carried out. The hub genes were screened using machine learning approaches, protein-protein interaction (PPI) network analysis, and Molecular Complex Detection (MCODE). A nomogram was developed for diagnostic evaluation. Furthermore, the relationship between hub gene expression profiles and immune infiltration landscapes was interrogated through immune cell infiltration analysis. The expression patterns of the hub genes were further validated in the II/R injury model through dataset validation and qRT-PCR assays. The procedure concluded in a gene-related hub network, DSigDB prediction of prospective therapeutic compounds, and molecular docking simulations of the drugs' binding affinity with important target proteins. Results: Hub genes have been found in five different DEMRGs: Pdk4, Yrdc, Bcl2l11, Bcl2a1d and Pmaip1. The nomogram model was beneficial for diagnosis. Dendritic cells (DC) and M2 macrophages are strongly linked to the 5 Hub genes, according to immune cell infiltration research. Afterwards, the regulatory network showed that hub genes and miRNAs had a complicated connection. Additionally, securinine and ABT-737 were anticipated to be possible therapeutic agents for II/R injury. The validation results for the four hub genes (Pdk4, Yrdc, Bcl2l11, and Pmaip1), obtained from both independent datasets and qRT-PCR, were consistent with the initial bioinformatics analysis. Conclusion: Pdk4, Yrdc, Bcl2l11, and Pmaip1 have been identified as hub genes closely associated with mitochondrial function in eraly II/R injury, thereby providing a theoretical basis for the diagnosis and treatment of eraly II/R injury.