Identification of Shared Genetic Biomarkers to Discover Candidate Drugs for Cervical and Endometrial Cancer by Using the Integrated Bioinformatics Approaches

Abstract

Cervical (CC) and endometrial cancers (EC) are two common types of gynecological tumors that threaten the health of females worldwide. Since their underlying mechanisms and associations remain unclear, computational bioinformatics analysis is required. In the present study, bioinformatics methods were used to screen for key candidate genes, their functions and pathways, and drug agents associated with CC and EC, aiming to reveal the possible molecular-level mechanisms. Four publicly available microarray datasets of CC and EC from the Gene Expression Omnibus database were downloaded, and 72 differentially expressed genes (DEGs) were selected through integrated analysis. Then, we performed the protein-protein interaction (PPI) analysis and identified 9 shared genetic biomarkers (SGBs). The GO functional and KEGG pathway enrichment analyses of these SGBs revealed some important functions and signaling pathways significantly associated with CC and EC. The interaction network analysis identified four transcription factors (TFs) and two miRNAs as key transcriptional and post-transcriptional regulators of SGBs. The expression of the AURKA, TOP2A, and UBE2C genes was higher in CC and EC tissues than in normal samples, and this gene expression was linked to disease progression. Furthermore, we performed docking analysis between 9 SGBs-based proteins and 145 meta-drugs, and identified the top-ranked 10 drugs as candidate drugs. Finally, we investigated the binding stability of the top-ranked three drugs (Sorafenib, Paclitaxel, Sunitinib) using 100 ns MD-based MM-PBSA simulations with UBE2C, AURKA, and TOP2A proteins, and observed their stable performance. Therefore, the proposed drugs might play a vital role in the treatment against CC and EC.