Identifying MicroRNA Biomarkers for Human Papillomavirus-infected Cervical Cancer Based on High-risk HPV Subtypes

• Main Authors: Po-Che Huang, 黃柏哲
• Other Authors: Tzong-Yi Lee
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/6h5ufh

碩士 === 元智大學 === 資訊工程學系 === 106 === Cervical cancer is the fourth most common cancer in women. More than 90 % cervical cancer cases are caused by human papilloma virus (HPV). There are several studies about identifying biomarkers of HPV-infected cervical cancer. However, those studies only focused on HPV16 or HPV18 infected cervical cancer, but in fact all of high-risk HPV are potential to lead cervical cancer. Therefore, we aim to discover the divergence between all high-risk HPV types infected cervical cancer to understand the biological mechanism by microRNA and gene expression data. We collected microRNA and gene expression data of HPV-infected cervical cancer from TCGA. According to HPV type of samples, we cluster into several subtypes and totally obtained six HPV subtypes including 16, 18, 45, 31, 33 and 39. Based on tumor samples of six HPV subtypes, we identified type-specific microRNAs of each HPV subtype. We also applied hierarchical method and principal components analysis to cluster HPV-infected cervical cancer samples by type-specific microRNAs. In addition, we performed machine learning approach such as correlation based feature selection and support vector machine (SVM) method to classify HPV-infected cervical cancer samples by type-specific microRNAs. Importantly, we have clinical NGS data to validate our type-specific microRNAs. Consequently, functional enrichment analysis (FEA) of type-specific microRNA-mediated genes will be performed to understand the biological mechanism of cervical cancer infected by different HPV subtypes. From functional enrichment analysis of type-specific microRNA-mediated genes, we discovered that there are some microRNA involve in biological process or pathway about HPV infection and cervical cancer. Importantly, some of them are also validated in our clinical NGS data. Based on those microRNAs, we construct the type-specific microRNA-mediated regulatory network of HPV 16, 18 and 33 to suggest a possible and significant role in HPV infection and cervical cancer. Especially in HPV18-specific microRNA-mediated regulatory network, hsa-mir-15b has been discovered that it will be induced with E2F-controlled genes which involved in differentiation, development, cell proliferation and apoptosis due to HPV infection.