Identifying COVID‐19 subtypes by single‐sample gene set enrichemnt analysis and providing guidance for sensitive drug selection

Abstract

AbstractThis study aimed at using single‐sample gene set enrichment analysis scores to cluster naso/pharyngeal swab specimen samples from coronavirus disease 2019 (COVID‐19) patients into two clusters. One cluster with higher fractions of immune cells and more active inflammatory‐related pathways was called the Immunity‐High (Immunity‐H) group, and the other one was called the Immunity‐Low group. We explored impacts of the method on COVID‐19 treatment. First, given that the Immunity‐H group was mainly enriched in inflammatory‐related pathways and had higher fractions of inflammatory cells, the Immunity‐H group may obtain more curative effects from anti‐inflammatory treatment. Second, we searched some hot genes from the PubMed platform that had been studied by researchers and found these genes upregulated in the Immunity‐H group, so we speculated the Immunity‐H group and Immunity‐Low group may have different curative effects from drugs targeting these genes. Finally, we screened out hub genes for the Immunity‐H group and predicted potential drugs for these hub genes by a public data set (http://dgidb.genome.wustl.edu). These hub genes are significantly upregulated in the Immunity‐H group and neutrophils so that the Immunity‐H group may obtain different treatment results from potential drugs compared with the Immunity‐Low group. Therefore, the cluster method may provide help in drug development and administration for COVID‐19 patients.