DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Yi, Gwan-Su | - |
dc.contributor.advisor | 이관수 | - |
dc.contributor.author | Kim, Yul | - |
dc.date.accessioned | 2019-08-22T02:42:04Z | - |
dc.date.available | 2019-08-22T02:42:04Z | - |
dc.date.issued | 2019 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=842080&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/264693 | - |
dc.description | 학위논문(박사) - 한국과학기술원 : 바이오및뇌공학과, 2019.2,[v, 60 p. :] | - |
dc.description.abstract | Discovering disease-associated genes becomes crucial issues for understanding disease mechanisms, diagnosis and therapy. Although the advance of high-throughput genomic and transcriptomic technologies helps re-searchers to identify novel disease-gene associations, these experimental methods are generally time and re-source consuming task. It also does not provide sufficient information and a number of disease-associated genes are still in a veil. Therefore, computational approaches have grown in importance to investigate candi-date genes more efficiently with low cost. In this paper, we developed a novel disease gene prioritization system, called DFENS-G, based on function similarity and network propagation. DFENS-G measured disease-gene as-sociations based on function module that are enriched with known disease genes, and propagated the infor-mation through an integrated network to identify novel candidates. We proved our system is superior to recon-struct known disease gene information compared with previous method using cross validation. Based on DFENS-G, we also identified genetic biomarkers for progression of rheumatoid arthritis and response of TNFa inhibitors. First, the multivariate logistic regression model was used to investigate the statistically significant SNPs. After that, to remove false positive markers, we filtered these SNPs based on their biologically related genes using the functional regions and eQTL information. We constructed prediction models by using support vector machine and validate the prediction performance with cross validation. As a result, the average accura-cy of our markers were superior than markers selected by statistically significance or a previous filtering meth-od. In conclusion, DFENS-G system can be directly used for identification and prioritization of reliable bi-omarkers through an insight of biological mechanism. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | function module similarity▼anetwork propagation▼agene prioritization▼arheumatoid arthritis▼abiomarkers▼adata mining▼amachine learning | - |
dc.subject | 기능 모듈 유사성▼a네트워크 전파▼a유전자 순위 화▼a류마티스관절염▼a바이오마커▼a데이터 마이닝▼a기계 학습 | - |
dc.title | Identification of novel biomarkers using recursive propagation with function module for predicting RA progression and drug response | - |
dc.title.alternative | 기능 모듈 반영 네트워크 전파 기반 류마티스관절염 예후 및 약물반응성 예측 바이오마커 발굴 | - |
dc.type | Thesis(Ph.D) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :바이오및뇌공학과, | - |
dc.contributor.alternativeauthor | 김율 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.