Design And Implementation Of DNA Methylation Status Online Prediction Platform

Epigenetics is one of the frontier topics of molecular biology in the post-genomic era. It aims to understand the heritable gene regulation that is not directly encoded in DNA sequence. DNA methylation is a main categories of epigenetic alterations, which played a important role in X-chromosome inactivation, genomic imprinting, ageing and tumorigenesis through gene expression regulation. In this dissertation, computational approaches are used to explore the methylation pattern of Cp G loci in the human genome, develop prediction model for predicting the methylation level of Cp G loci and DNA methylation status online prediction platform. When used the platform study on DNA methylation characteristic of rheumatoid arthritis, we got some interesting findings.The main contributions are as follows:First, a model predicting the methylation level of Cp G loci is developed. Besides the whole genome bisulfite sequencing DNA sequence features Infinium Human Methylation 450 Bead Chip data are added in the classifier construction of the model. This model not only shows better prediction results compared with those using only DNA sequence features, but also quantifies the correlation between the methylation level of a Cp G site and its adjacent upstream and downstream Cp G sites in the same cell type or tissue.Second,a DNA methylation online prediction platform base on the study of methylation level prediction model is Designed and implemented. In this dissertation, a DNA methylation online prediction platform based on J2 EE and R operation engine is implemented. The platform with good scalability implemented by B/S architecture and modular design.Third, study on characteristic of rheumatoid arthritis used DNA methylation online prediction platform. The results show that the DNA methylation online prediction platform significantly expands the coverage of Illumina 450 K chip detection, which combined Omics, it can better identify genes and pathways are closely related to the pathogenesis of RA.In summary, the prediction mode compared to the conventional method has higher accuracy in prediction, the prediction platform in the study of methylation characteristic of rheumatoid arthritis show that the platform has a certain versatility.