Prediction of the risk of non-viral hepatocarcinogenesis by machine learning using genomic information of host and intestinal commensal bacteria

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K08288
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 53010:Gastroenterology-related
• Research Institution: Hiroshima University
• Principal Investigator: Miki Daiki 広島大学, 医系科学研究科(医), 講師 (10584592)
• Project Period (FY): 2020-04-01 – 2023-03-31
• Keywords: ヒトゲノム / 腸内細菌 / 非ウイルス性肝癌 / 機械学習

Outline of Final Research Achievements

Non-viral liver cancers are on the rise. One of the causes is the increase in non-alcoholic steatohepatitis (NASH). On the other hand, it has become clear that the gut microbiota is involved in various diseases including diabetes, obesity, and even liver diseases. In this study, we aimed to extract non-viral hepatocarcinogenic risk factors by integrating the human genome, its transcripts, and the genomes of commensal microorganisms, i.e., intestinal bacteria, using machine learning methods. As a result, we were able to construct several models for predicting hepatocarcinogenesis. We plan to prospectively confirm the prediction accuracy of each model.

Free Research Field

ゲノム医科学、肝臓学

Academic Significance and Societal Importance of the Research Achievements

これまで経験を積み上げてきたヒトゲノムあるいはがんゲノムやトランスクリプト研究といった特徴的な基盤の上に、さらにメタゲノムという共生する非宿主生物由来の新たな情報を加えることで、独自のアプローチで非ウイルス性肝癌のリスク因子に迫ることを目指す。食生活の変化を含めた環境要因などの外的要因がリスクを高めるとも推測されるが、それに影響をうけて可逆的に変化をするメタゲノム情報と、一生変わることなく安定なヒトゲノム情報とを網羅的に統合解析するのは、新たな手法である。これまでの単階層解析では見えてこなかった未知の特徴量の抽出を行い、より有用な疾患リスク評価を行うことで、これら疾患群の診療への貢献を目指す。