Predict disease-related RNA methylation sites from the methylation-expression association by using hypergeometric test

Conference: BIBE 2019 - The Third International Conference on Biological Information and Biomedical Engineering
06/20/2019 - 06/22/2019 at Hangzhou, China

Proceedings: BIBE 2019

Pages: 7Language: englishTyp: PDF

Personal VDE Members are entitled to a 10% discount on this title

Tang, Yujiao; Chen, Kunqi; Wei, Zhen; Wu, Xiangyu; Meng, Jia (Department of Biological Sciences, Research Center for Precision Medicine, Xi’an Jiaotong-Liverpool University, Suzhou, China)

N6-methyladenosine (m6A) is the most abundant RNA modification on mRNA and lncRNA in human. Recent studies have shown that it is implicated in various critical biological processes, such as translation and alternative splicing, and involves in multiple human diseases, including cancer and obesity. However, only a small number of RNA methylation sites have been explicitly associated to disease conditions with experimental approaches, since RNA m6A methylation sites may potentially play a pivotal regulatory role in a wide range of human pathogenesis and should not be ignored, an efficient predictor for disease-associated m6A RNA methylation sites becomes a major challenge. In order to obtain a comprehensive understanding of disease-associated m6A RNA methylation site, we purpose here a computational framework to integrate the three different layers of a network structure, including the expression profiles of genes, the methylation profiles of m6A RNA methylation sites and gene-disease associations, and utilize subsequently the Hypergenomic test to predict integrally the potential disease-associated m6A sites. We show with a rigorous cross-validation that the AUROC of the proposed approach is 0.73; and a number of predictions are supported by existing literatures, suggesting our prediction is helpful for identifying the novel m6A sites associated to human disease. Ultimately, the predicted results are freely available online at:, which supports the queries of diseaserelated m6A RNA methylation sites. We presented a very first attempt for computational prediction of diseaseassociated RNA methylation sites, helping researchers of the field to understand the roles of m6A RNA methylation in human diseases and facilitating the development of the treatments.