Objective:To conduct data mining based on bioinformatics databases to explore key genes related to the occurrence of childhood systemic lupus erythematosus(cSLEs)and build a risk score model.Methods:cSLE-related data were obtained from the Gene Expression Comprehensive Database.Genes related to the onset of cSLE were identified through various machine learning methods and a risk score model was constructed.TNFAIP6 expression was verified by ELISA,and the effect of TNFAIP6 on cell apoptosis and cell cycle was analyzed by flow cytometry.Results:TNFAIP6,B4GALT5,HLX,ANXA3 and DYSF were related to the onset of cSLE and had strong diagnostic value for cSLE.Their areas under the curve were TNFAIP6:0.866,B4GALT5:0.891,HLX:0.914,ANXA3:0.878,DYSF:0.929.The risk score model constructed using TNFAIP6 and DYSF could effectively diag-nose cSLE(areas under the curve was 0.969).Neutrophil levels were significantly elevated in the high-risk group compared to the low-risk group(t=268.5,P=0.009).Experimental results showed that TNFAIP6 was highly expressed in the serum of cSLE patients,and its silencing might promote THP-1 cell apoptosis and arrest the cell cycle.Conclusion:The cSLE risk score model constructed us-ing TNFAIP6 and DYSF can effectively identify cSLE;TNFAIP6 may be a potential biomarker for cSLE.
维普
万方数据