[Construction and discussion of risk prediction model for allergic asthma in children].

A prediction model for the risk of childhood allergic asthma was established through the analysis of public datasets. By using bioinformatics analysis methods, two datasets, GSE40732 and GSE40888, were selected, which included the whole-genome expression profile data of 222 children. Among them, GSE40732 was used as the training dataset to detect differentially expressed genes in peripheral blood mononuclear cells of children with the disease, and the master regulator analysis (MRA) algorithm was used to screen the master regulator genes in the inflammation-related pathway (GO: 0006954). After obtaining the master regulator genes, the expression of these master regulator genes in the GSE40732 and GSE40888 datasets was detected, and a prediction model was constructed through logistic regression, based on which risk scores were assigned to children. By comparing the risk scores of healthy children and children with the disease, the area under the curve (AUC) was used to evaluate the classification performance of the model. The average value of the risk scores of all children with the disease output by the model was calculated as the threshold. According to this threshold, the children with the disease in the two datasets were divided into high-risk and low-risk groups. The CIBERSORT algorithm was applied to analyze the infiltration of immune cells in the high-risk and low-risk groups, and the enrichment analysis of signaling pathways was completed using the msigdbr package in R software. The results showed that compared with healthy children, there were 377 up-regulated genes and 255 down-regulated genes in the peripheral blood mono-nuclear cells of children with the disease. The MRA algorithm analysis showed that there were five genes (<i>MUC5B</i>, <i>CST4</i>, <i>CCR7</i>, <i>TNF-α</i>, and <i>THBS1</i>) that were the master regulator genes in the regulatory network. Risk score=<i>MUC5B</i>×3.47<i>+CST4</i>×2.17<i>+CCR7</i>×0.59<i>+TNF</i>-<i>α</i>×0.54<i>+THBS1</i>×1.67. The AUC in the GSE40732 and GSE40888 datasets were 0.874 and 0.682, respectively. Compared with the low-risk group, the resting memory CD4⁺T cells and regulatory T cells in the peripheral blood of children with the disease in the high-risk group significantly decreased (<i>P</i><0.05), and both the IL-33 and IL-13 pathways were highly enriched. In conclusion, the model constructed in this study has a good predictive efficiency for the risk of allergic asthma and also has a certain effect on risk stratification.