In developing countries, cervical cancer(CC) is the most second common cancer[1]. This cancer is a type of cancer which occurs in the cervix cells[2]. In another word, the growth of abnormal cells in the lining of the cervix is called CC. CC includes two common histological types: Squamous cell carcinoma (SCC) and adenocarcinoma (ADC)[3]. In this study, SCC has been studied. Squamous cell carcinoma is the most common CC, accounting for 70% of cases.
Cancer biomarker detection is one of the important challenges in cancer studies. Cancer biomarker is a biological molecule which shows the presence of cancer[4]. There are different types of cancer biomarkers including molecular, radiographic, histologic, and physiologic[5]. The goal of this study is to identifying molecular biomarkers for CC. The molecular samples which have analyzed in this study is transcriptome profiles.
Different studies have been done in order to discover RNA-based cancer biomarkers. Chen-Xia Wen et al.[6] in their research introduced miRNA-873-5p as potential biomarker and promising therapeutic approach for CC. Xiaoli Cao et al.[7] reported CCAT2 as candidate biomarker for diagnosis and prognostic prediction of CC. In other study, EFNA1 is introduced as a novel prognostic biomarker for CC by Xiaopeng Shen ad colleagues. As well as, Anja Nilsen and colleagues[8] proposed miR-200a, miR-200b, and miR-429 as candidate biomarkers in CC. Moreover, Kaidi Zhao et al. [9] concluded that SPP1 can be a Prognostic Biomarker in CC. In other study, INHBA was introduced as a prognostic biomarker and also reported that this gene is corelated with Immune Cell Infiltration in CC[10]. Besides, Xinyang Zhang and colleague introduced a Circular RNA named circYPEL2 as a potential biomarker for clinical research of CC.
Moreover, some network-based studied have been done for identifying cancer biomarkers. Nancy Lan Guo and colleagues[11], proposed a network-based algorithm for identification of cancer biomarkers. In 2021, Youwei Hua et al. introduced a gene co-expression network for identification of the biomarkers in human tumors[12]. In other study, Shuyan Zhang and colleague[13] proposed two lncRNA as prognostic biomarker in gastric cancer based on integrated analysis of lncRNA-Associated ceRNA network. A deep learning model and similarity network fusion was proposed by Tzu-Hao Wang and colleague[14] for identification of biomarker through multiomics data analysis in Prostate Cancer. Yanqiu Tong et al.[15] have done theorical and in silico analyses and proposed MYC as a dynamic network biomarker in colon and rectal cancer. As well as, Faping Li and colleague[16] proposed a competing endogenous network for identification of prognostic biomarker in bladder cancer. In 2020, Yuan Yang et al. introduced a gene regulation network analysis and proposed YAP1 as a prognostic biomarker in pancreatic cancer.
Proteins control biological processes, molecular functions and cellular mechanism and determine disease and healthy states[16]. Therefore, study of proteins' interactions inside the cell are very important. Thus, in the current project a Protein-Protein Interaction (PPI) network analysis was studied.
miRNAs are small non-coding RNA molecules which regulate mRNAs from being translated [17, 18]. This type of RNAs regulate gene expression at the posttranscriptional level and can be found in tissue, blood and body fluids[19]. Recently, miRNAs introduced as prognostic and diagnostic biomarkers in different cancers such as breast, colorectal, ovarian and cervical cancers. Interaction between miRNAs and genes show the regulatory relationship between miRNA and genes[20]. In this regard, miRNA-mRNA interaction networks have been studied in different cancer studies. Negar and colleagues[21] introduced a miRNA-mRNA network-based biomarker for Alzheimer disease. Beside, Habib at al.[22] proposed a miRNA-mRNA module prognostic biomarker for early detection of colorectal cancer based on co-expression network analysis. As well as, Masoumeh and colleagues[23] proposed a miRNA-mRNA sub-network as prognostic biomarker for breast cancer subtype stratification. In this project, interactions of miRNAs and target genes have been studied and two significant miRNAs introduces as prognostic biomarker for cervical cancer.
Cancer driver genes are the genes in which mutations in these genes cause to tumor growth[24]. These genes can be of two types: tumor suppressor genes and proto-oncogenes. DriverDBv3[25] is an online database which contains human cancer driver genes with Mutation, CNV, and Methylation information. In the current project, list of driver genes for cervical cancer was collected from this database.
The current study aimed to identify the genes and miRNAs as prognostic biomarkers in CC. In this regard, this project used a PPI network analysis to discover candidate prognostic biomarkers for CC. In this project, at first a transcriptome profile for normal and CC samples were downloaded from the NCBI-GEO with accession number GSE63514. Then, differentially expressed genes (DEGs) between normal and cervical cancer groups was calculated and list of significant genes was selected for network construction. Next, a PPI network was constructed for the selected genes in STRING[26] online tool. After that, a significant protein module was extracted from the PPI network. Subsequently, list of miRNAs which are targeting module's genes were collected from the miRTarBase[27] online database. Consequently, four driver genes (MCM2, MCM10, POLA1 and TONSL) along with two miRNAs (miR-193b-3p and miR-615-3p) were introduced as prognostic biomarker in CC. The workflow diagram of this project is depicted in Fig. 1.