The risk of COVID-19 can be predicted by a nomogram based on m6A-related genes

BACKGROUND: The expression of m6A-related genes and their significance in COVID-19 patients are still unknown. METHODS: The GSE177477 and GSE157103 datasets of the Gene Expression Omnibus were used to extract RNA-seq data. The expression of 26 m6A-related genes and immune cell infiltration in COVID-19 patients were analyzed. Finally, we built and validated a nomogram model to predict the risk of COVID-19 infection. RESULTS: There were significant differences in 11 m6A regulatory factors between patients with COVID-19 and healthy individuals. The classification of disease subtypes based on m6A-related gene levels can be distinguished. COVID-19 patients in GSE177477 were classified into two categories based on m6A-related genes. The patients in cluster A were all symptomatic, while those in cluster B were asymptomatic. A significant correlation was also found between immune cells and m6A-related genes. Finally, seven m6A-related disease-characteristic genes, HNRNPA2B1, ELAVL1, RBM15, RBM15B, YTHDC1, HNRNPC, and WTAP, were screened to construct a nomogram model for predicting risk. The calibration curve, decision curve analysis, and clinical impact curve analysis were used to show that the nomogram model was effective and had a high net efficacy for risk prediction. CONCLUSIONS: m6A-related genes were correlated with immune cells. The nomogram model effectively predicted COVID-19 risk. Moreover, m6A-related genes may be associated with the presence or absence of symptoms in COVID-19 patients.