상호정보와 부울대수를 이용한 복합질환의 SNP 상호작용 예측 = Prediction of SNP interactions in complex diseases with mutual information and boolean algebra

국문 초록 (Abstract)

대부분의 만성질환은 다수의 유전자-유전자 사이의 상호작용에 의해서 발병하는 복합질환이다. 복합질환의 발병에 관여하는 단일염기다형성(single nucleotide polymorphism; SNP)과 유전자 사이의 상호작용을 찾아내는 연구는 질환의 예방과 치료에 기여한다. 기존의 연구 방법은 주로 특정 유전자 내 SNP 조합을 찾아내는 데 그치고 있다. 본 연구에서는 SNP 조합의 구성원 사이에 일어나는 구체적인 상호작용을 나타내는 부울식을 찾는 방법을 제시한다. 본 논문에서 제안하는 방법은 두 단계로 이루어진다. 제 1 단계에서는 엔트로피에 기반한 상호정보를 이용하여 발병에 관여하는 SNP 조합을 찾는다. 제 2단계에서는 제 1 단계에서 찾은 SNP 조합으로 이루어지는 부울식 중에서 발병 예측정확도가 가장 높은 부울식을 찾는다. 제안한 방법을 임상자료에 적용하여 그 효율성을 실험하였으며 기존 연구들과 장단점을 비교하였다.

다국어 초록 (Multilingual Abstract)

Most chronic diseases are complex diseases which are caused by interactions of several genes. Studies on finding SNPs and gene-gene interactions involved in the development of complex diseases can contribute to prevention and treatment of the diseases. Previous studies mostly concentrate on finding only the set of SNPs involved. In this study we suggest a way to see how these SNPs interact using boolean expressions. The proposed method consists of two stages. In the first stage we find the set of SNPs involved in the development of diseases using mutual information based on entropy. In the second stage we find the highest accuracy boolean expression that consists of the SNP set obtained in the first stage. We experimented with clinical data to demonstrate the effectiveness of the proposed method. We also compared the differences between our method and the previous results on the SNP associations studies.

참고문헌 (Reference)

1 신진섭, "생체정보측정을 통한 진단시스템 개발" 한국컴퓨터정보학회 13 (13): 219-226, 2008

2 이재원, "생명과학 연구를 위한 통계적 방법" 자유 아카데미 2005

3 김광백, "개선된 퍼지 ART 알고리즘을 이용한 한방 자가 진단 시스템" 한국컴퓨터정보학회 15 (15): 27-34, 2010

4 A.Litonjua, "The significance of beta2-adrenergic receptor polymorphisms in asthma" 12 (12): 12-17, 2006

5 C.Kooperberg, "Sequence analysis using logic regression" 21 : 626-631, 2001

6 C. Tasi, "Renin-angiotensin system gene polymorphisms and atrial fibrillation" 109 (109): 1640-1646, 2004

7 X.Wan, "Predictive rule inference for epistatic interaction detection in genome-wide association studies" 26 (26): 30-37, 2010

8 J.Moore,, "Power ofmultifactor dimensionality reduction fordetecting gene-gene interactions in the presence of genotyping error,missing data,phenocopy,and genetic heterogeneity" 24 (24): 150-157, 2003

9 S. Kim, "Polymorphism of tandem repeat in promoter of 5-lipoxygenase in ASA-intolerant asthma: a positive association with airway hyperresponsiveness" 60 (60): 760-765, 2005

10 M. Ritchie, "Multifactordimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer" 69 (69): 138-147, 2001

11 D. Contopoulos-Ioannidis, "Meta-analysis of the association of beta2-adrenergic receptor polymorphisms with asthma phenotypes" 115 (115): 963-972, 2005

12 X.Wan,, "MegaSNPHunter:a learning approach to detect disease prediction SNPs and high level interactions in genome wide association study" 10 (10): 2009

13 A.Tarca, "Machine Learning and Its Applications to Biology" 3 (3): 953-963, 2007

14 T.Cormen, "Introduction to algorithms" MIT Press 2009

15 C.Kooperberg, "Identifying interacting SNPs using Monte Carlo logistic regression" 28 (28): 157-170, 2005

16 K. Fukunaga, "Genetic polymorphisms of CC chemokine receptor 3 in Japanese and British asthmatics" 17 : 59-46, 2001

17 T.Cover, "Elements of information theory" Wiley 2006

18 R.Culverhouse,, "Detecting epistatic interactions contributing to quantitative traits" 27 (27): 141-152, 2004

19 S. Kim, "Cysteinyl leukotriene receptor 1 promoter polymorphism is associated with aspirin-intolerant asthma in males" 36 (36): 433-439, 2006

20 J.Moore,, "Bioinformatics challenges for genome-wide association studies" 26 (26): 445-455, 2010

21 Y.Zhang, "Bayesian inference of epistatic interactions in case-control studies" 39 (39): 1167-1173, 2007

22 K.Liang, "An algorithm for model construction and its napplications to pharmacogenomic studies" 51 : 751-759, 2006

23 X.Chen, "A forest-based approach to identifying gene and gene-gene interactions" 104 (104): 19199-19203, 2007

24 M.Nelson,, "A combinatorial partitioning method to identify multilocus genotypic partitions that predict quantitative trait variation" 11 : 458-470, 2001