Efficient knowledge discovery from life-science big data

2017 Fiscal Year Final Research Report

• Project/Area Number: 26330334
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Life / Health / Medical informatics
• Research Institution: Nagahama Institute of Bio-Science and Technology
• Principal Investigator: Ikemura Toshimichi 長浜バイオ大学, バイオサイエンス学部, 客員教授 (50025475)
• Co-Investigator(Kenkyū-buntansha): 和田 健之介 長浜バイオ大学, バイオサイエンス学部, 教授 (90231026)
• Project Period (FY): 2014-04-01 – 2018-03-31
• Keywords: ビッグデータ / 人工知能 / 時系列解析 / メタゲノム解析 / 核酸医薬 / 自己組織化マップ / 感染症ウイルス / オリゴヌクレイチド

Outline of Final Research Achievements

We have developed a BLSOM (Batch-learning SOM) for oligonucleotide compositions, which is suitable for big data analysis of genome sequences. It is an unsupervised machine-learning and thus enables unexpected knowledge discoveries. By using the AI-method for genomes of RNA viruses such as influenza virus, ebolavirus and MERS coronavirus, we found time-series and reproducible changes in oligonucleotide composition and, for these highly mutable viruses, we have developed a method for designing oligonucleotide drugs with long-term efficacy. By BLSOM analyses on higher animals such as human and Coelacanth, we found regions enriched by specific 5-mers that include CpG, a representative epigenetic marker, or by binding sequences of various transcription factors. These specific regions cluster densely in the vicinity of centromeres, which are known to form chromocenters. Therefore, we proposed a model on their roles in nuclear organization.

Free Research Field

ゲノム進化の情報解析