GENOME-WIDE ANALYSIS OF HBV INTEGRATION SITES IN CANCER: FROM TOOL DEVELOPMENT TO THE DISCOVERY OF POTENTIAL MECHANISMS

Abstract

One of the most important strategies in HBV-inducing hepatocellular carcinoma (HCC) is HBV integration, during which a fragment of the virus genome is inserted into the infected host‟s genome. First, we propose a computational pipeline, VirusInsert,to detect virus integration sites using NGS data. Second, we compared the performance of VirusInsert with two other state-of-the-art virus integration tools on a public dataset, in-house dataset, and our simulation dataset. The experimental results have shown the superiority of our tool in both its detection accuracy and sensitivity. Last, we utilized large-scale clinical samples from TCGA database to validate the functional potential of the detected integration sites by VirusInsert. Remarkably, several recurrent integrated genes and functional pathways were identified using the detected integration sites. In conclusion, these findings support the biological contribution of the proposed tool and reveal the potential role of HBV integration in HCC induction and progression.