Machine learning-based analysis of diagnostic biomarkers for biliary atresia and their relationship with immune cell infiltration
Objective:To screen diagnostic biomarkers for biliary atresia and their correlation with immune cell infiltration based on bioinformatics technology.Methods:The gene expression profiles related to biliary atresia were obtained from gene expression omnibus(GEO)database,and the differential genes related to biliary atresia were screened by R language,and gene enrichment analysis was conducted.The main module genes selected by differential gene and weighted gene co-expression network were collected,core genes were identified by using two machine learning algorithms:LASSO regression and SVM-RFE analysis.Receiver operating characteristic(ROC)curve was used to verify the diagnostic accuracy of core genes for biliary atresia.The infiltration levels of 28 kinds of immune cells were detected by ssGSEA and their association with core genes was analyzed.Results:A total of 76 differential genes were screened.Enrichment analysis showed that differential genes were mainly enriched in extracellular matrix receptor interaction and interleukin-17(IL-7)signaling pathway and phosphatidylinositol 3-kinase/protein kinase B(PI3K/Akt)signaling pathway.After overlapping of key modules and differential genes,26 overlapping genes were screened.Two machine learning algorithms,SVM-RFE and Lasso,were used to identify three core genes(CXCL8,LAMC2,KRT19)as potential diagnostic biomarkers for biliary atresia.Differential analysis showed that the expression levels of CXCL8,LAMC2 and KRT19 in children with biliary atresia were significantly higher than those in normal controls(P＜0.001).ROC curve analysis showed that CXCL8,LAMC2,and KRT19 had high diagnostic value,and the area under ROC curve of the combined diagnosis was 0.969.It had been proven to be positively correlated with concentrated immune cell infiltration.Conclusions:CXCL8,LAMC2,and KRT19 are all up-regulated in the liver tissue of children with biliary atresia,and their combined diagnosis has high accuracy.They may be candidate genes for the diagnosis of biliary atresia,and provide new ideas for subsequent treatment.
万方数据
维普
