Database screening of herbal monomers regulating autophagy by constructing a "disease-gene-drug" network

BACKGROUND Studies suggest an important role of autophagy as a target for cancer therapy. We constructed a "disease-gene-drug" network using the modular approach of bioinformatics and screened herbal monomers demonstrating functions related to autophagy regulation. METHODS Based on the microarray results of the gene expression omnibus (GEO) database (GSE2435 and GSE31040, starvation-induced autophagy model), we used the human protein reference database (HPRD) to obtain the protein-protein interaction (PPI) network. In addition, we used the CFinder software to identify several functional modules, performed gene ontology-biological process (GO-BP) functional enrichment analysis using the DAVID software, constructed a herbal monomer-module gene regulatory network using literature search and the Cytoscape software, and then analyzed the relationships between autophagy, genes, and herbal monomers. RESULTS We screened 544 differentially expressed genes related to autophagy, 375 pairs of differentially expressed genes, and 7 gene modules, wherein the functions of module 3 (composed of 7 genes) were enriched in "cell death". Using the constructed herbal monomer-module gene regulatory network, we found that 30 herbal monomers can simultaneously regulate these 7 genes, indicating a potential regulatory role in autophagy. CONCLUSIONS Database screening using the disease-gene-drug network can provide new strategies and ideas for the application of herbal medicines in cancer therapy.