Screening of the co-diagnostic gene MMP9 for ischemic stroke and obesity based on bioinformatics analysis and machine learning

Objective To explore the pathogenesis of obesity (OB) and ischemic stroke (IS), common diagnostic genes, genomic enrichment analysis and genomic variation analysis by using bioinformatics and machine learning methods, and to find the correlation between common diagnostic genes and immune cells.Methods OB and IS related chips were obtained from GEO database, R language was used to analyze gene differences and WGCNA analysis, and GO and KEGG enrichment analysis was performed at the intersection. PPI network was constructed at the same time. Three machine learning and 12 cytohubba methods were used to screen key genes, ROC curve and sample chip were used to select the best co-diagnostic genes, and CIBERSORT algorithm was continued to analyze the immune infiltration of OB and IS.Results A total of 235 differential genes were obtained in GSE25401 and GSE151839 in OB training group, and 525 differential genes were obtained in GSE22255 and GSE37587 in IS training group. GO analysis was mainly involved in the regulation of myeloid leukocyte differentiation. KEGG analysis showed that the pathways of complement and coagulation cascade were significant. Three machine learning methods and 12 cytohubba methods were used to select the key genes MMP9 and TNFAIP6. ROC curve line and sample chip verification showed that MMP9 had the best effect. Finally, MMP9 was used for genome enrichment analysis and genome variation analysis.Conclusion OB and IS are common diagnostic genes, MMP9 is the future fertilizer the diagnosis and treatment of obesity and ischemic stroke provide potential targets and value.