초고성능컴퓨팅 기반 건강한 고령사회 대응 빅데이터 기술개발

Abstract

본 연구과제의 목적은 HPC(High Performance Computing) 및 데이터 기반의 질병 연구를 위한 분석 기술의 개발임. 이를 위해, 본 연구과제는 당해 연도에 다음과 같이 수행되었음.

○ 노인성 치매 대응을 위한 치매 데이터 수집 및 네트워크 구축 - 생물학적 데이터 수집 및 네트워크 구축 - 치매 원인 규명을 위한 유전체분석 기술 ○ 생물학적 네트워크 분석기술 연구개발 - 생물학적 네트워크 분석(유사매칭기술) - 네트워크 분석기반의 바이오 마커 및 관련 핵심 화합물 발굴기술 ○ 치매 연구 지원 시스템 개발 및 활용연구 지원

(출처 : 보고서 초록 5p)

Ⅳ. Results of the research projects ○ Significance of data network analysis for dementia - Development of a biological network similarity matching algorithm that has 160% faster performance compared to a previous SAGA method - In-vitro experimental validation of in-silico candidate compounds that interact with target protiens - Publication of four research papers on the top 20% international journals

○ Usefulness of data network analysis for dementia - Improvement of entity name recognition technology by deep learning models (CNN LSTM CRF) - Development of a genomic data pre-processing system for big data analysis based on multi-node execution environment (8.41 times faster than single-node execution environment) - Identification of brain-specific 1,287 somatic variants in dementia patients (including more than two variants that are highly associated with dementia pathogenesis) - Completion of trio whole genome data analysis using trio data of 201 autism families and identification of genetic variants

○ Development and application of a research support prototype system - Development of an expression/network analysis platform (AlzNAVi) for dementia based on graph database that integrate nine public DBs - User satisfaction evaluation for the AlzNAVi (the average score is 88.1 from 64 users) - Identification of 15 candidate genes for Alzheimer disease based on big data analysis and evaluation of 4 genes that induce hyperphosphorylation of Tau protein

(출처 : SUMMARY 10p)