Identification of Significant Modules and Enriched GO and KEGG Pathway Analysis

First, the correlation between module eigengenes (MES) and clinical features was calculated to determine the relevant modules. In principal component analysis, gene significance (GS) is the log10 conversion of the P value in linear regression between gene expression and clinical information (GS=LG P). MES were taken as the principal components of each gene module, and the expression patterns of all genes were summarized into a single characteristic expression profile within a specific module. The correlation between MES and disease state (SLE vs control) was then calculated to determine the relevant modules. Leveraging online annotation, visualization, and integrated discovery databases (David; https://david.ncifcrf.gov/summary.jsp) [19,20], the biological progress (BP), cell components (CC), molecular function (MF), and the Kyoto Encyclopedia Gene and Genome (KEGG) pathways can be analyzed, furthering the understanding of gene function in the key module. The R package “VennDiagram” plots a Venn diagram by overlapping the key modules with the genes in the ARGs. Statistical significance was set at P<0.05. Gene annotation information on the platform was used to transform the probes into homologous gene symbols. Next, a total of 799 ARGs were obtained by searching the Human Autophagy Database (http://autophagy.lu/). The recursive feature elimination algorithm can improve the classification accuracy and select risk genes from feature ARGs [21,22].