TY - GEN
T1 - A Markov random field-based Bayesian model to identify genes with differential methylation
AU - Wang, Xiao
AU - Gu, Jinghua
AU - Xuan, Jianhua
AU - Clarke, Robert
AU - Hilakivi-Clarke, Leena
PY - 2014
Y1 - 2014
N2 - The rapid development of biotechnology makes it possible to explore genome-wide DNA methylation mapping which has been demonstrated to be related to diseases including cancer. However, it also posts substantial challenges in identifying biologically meaningful methylation pattern changes. Several algorithms have been proposed to detect differential methylation events, such as differentially methylated CpG sites and differentially methylated regions. However, the intrinsic dependency of the CpG sites in a neighboring area has not yet been fully considered. In this paper, we propose a novel method for the identification of differentially methylated genes in a Markov random field-based Bayesian framework. Specifically, we use Markov random field to model the dependency of the neighboring CpG sites, and then estimate the differential methylation score of the CpG sites in a Bayesian framework through a sampling scheme. Finally, the differential methylation statuses of the genes are determined by the estimated scores of the involved CpG sites. In addition, significance test is conducted to assess the significance of the identified differentially methylated genes. Experimental results on both synthetic data and real data demonstrate the effectiveness of the proposed method in identifying genes with differential methylation patterns under different conditions.
AB - The rapid development of biotechnology makes it possible to explore genome-wide DNA methylation mapping which has been demonstrated to be related to diseases including cancer. However, it also posts substantial challenges in identifying biologically meaningful methylation pattern changes. Several algorithms have been proposed to detect differential methylation events, such as differentially methylated CpG sites and differentially methylated regions. However, the intrinsic dependency of the CpG sites in a neighboring area has not yet been fully considered. In this paper, we propose a novel method for the identification of differentially methylated genes in a Markov random field-based Bayesian framework. Specifically, we use Markov random field to model the dependency of the neighboring CpG sites, and then estimate the differential methylation score of the CpG sites in a Bayesian framework through a sampling scheme. Finally, the differential methylation statuses of the genes are determined by the estimated scores of the involved CpG sites. In addition, significance test is conducted to assess the significance of the identified differentially methylated genes. Experimental results on both synthetic data and real data demonstrate the effectiveness of the proposed method in identifying genes with differential methylation patterns under different conditions.
KW - Bayesian framework
KW - Gibbs sampling
KW - Markov random field
KW - dependency structure
KW - differential methylation events
KW - significance test
UR - http://www.scopus.com/inward/record.url?scp=84904430396&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84904430396&partnerID=8YFLogxK
U2 - 10.1109/CIBCB.2014.6845515
DO - 10.1109/CIBCB.2014.6845515
M3 - Conference contribution
AN - SCOPUS:84904430396
SN - 9781479945368
T3 - 2014 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology, CIBCB 2014
BT - 2014 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology, CIBCB 2014
PB - IEEE Computer Society
T2 - 2014 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology, CIBCB 2014
Y2 - 21 May 2014 through 24 May 2014
ER -