Discovery and Identification of Candidate Genes, Diagnostic Model and Drug Predictions for Schizophrenia and Crohn's Disease Through Integrated Bioinformatics Analysis and Machine Learning

Abstract

Abstract Background Both schizophrenia and Crohn's disease are linked to the expression of immunological and metabolic abnormalities. The objective of this research is to find appropriate diagnostic candidate genes for patients with schizophrenia and Crohn's disease as well as the appropriate medications. Methodology: The datasets were retrieved from the Gene Expression Omnibus (GEO) database for schizophrenia and Crohn's disease. Differentially expressed genes (DEGs) were identified using the Limma package and weighted gene co-expression network analysis (WGCNA). The function enrichment analysis was conducted, followed by machine learning-based identification of candidate immune-related central genes for least absolute shrinkage and selection operators (LASSO) regression and construction of protein-protein interaction (PPI) network. The random forest method was used to identify candidate genes and establish artificial neural networks for the verification of these genes. And for the diagnosis of schizophrenia, the receiver operating characteristic (ROC) curve was plotted. The Enrichr database was then used to collect pertinent drugs that are related to candidate genes. Results A total of 2681 DEGs and 210 Crohn's disease-related genes were screened out in schizophrenia. Among the schizophrenia and Crohn's disease-related genes, about 35 genes depicted differential expression. Finally, seven potential genes were screened out using the PPI network and machine learning. The diagnostic value was evaluated using the ROC curve data. These findings suggest that the diagnostic value of these candidate genes is high. Valproic acid and other related drugs were collected from the enrichr database. Conclusion The identification of seven candidate gene models (CAP1, INSIG1, MSMO1, PHLDA2, PSMB6, TBC1D2, UBA5) has high diagnostic value, and valproic acid and other drugs may become candidate drugs for patients with schizophrenia, providing effective evidence for the pathogenesis and treatment.