Font Size: a A A

Identification Of Characteristic DNA Methylation Markers In Human Genome

Posted on:2017-04-21Degree:DoctorType:Dissertation
Country:ChinaCandidate:F WangFull Text:PDF
GTID:1360330596958391Subject:Biomedical engineering
Abstract/Summary:PDF Full Text Request
Gene expression is an important step that can transformate genetic information to protein with activity and function during the life of organism.In this process,the changes of gene expression may cause the abnormal of protein function,cell differential,tissue formation even to tumorigenesis.It has known that DNA mthylation as an epigenetic modification play an important role in regulation of gene expression,involved in cell differential and tumorgenesis.Therefore,identification as well as functional analysis of DNA methylation marker has been a hot of epigenetic research.However,DNA methylation especially DNA methylation pattern shows high divergence between species,individuals,even to tissues.We refined the DNA methyaltion markers from species to individual to tissue,which may provide a theoretical basis to explore the epigenetic regulatory mechanism.Using DNA methylation data of primate species including human,chimpanzee and rhesus,we filtered the variation of intra-species DNA methylation between individuals or Cp Gs through integration of mixed effect model and residual maximum likelihood model.The methylation level of each species was precise quantitative to identify the human specific DNA methylation markers.These DNA methylation markers enriched in hypomethylation of promoter,negatively regulated gene expression,involved in development process and specially enriched in neuronal related transcript factor.The DNA methylation markers specially selectived in human have high conserved DNA sequence.It is suggested that the highly conserved DNA sequence may rapidly evolute in epigenetic,involved in the formation of species-specific phenotype and function.Abnormal DNA methylation may induce the transcriptional silencing,leading to differences of tumorgenesis,development and prognosis.However,there are only weak correlation between average methylation in promoter and gene expression.We re-evaluate the regulatory effects of Cp Gs in promoter on gene expression to re-quantative the DNA methylation level in DNA methylation data of breast cancer through canonical correlation analysis,and showed higher correlation with gene expression than average methylation.Based on the new methylation levels of promoters,we idenitify the DNA methylation markers associated with the survival time of breast cancers.The DNA methylation markers not only significant distinguish the survival time of breats cancers but also showed higher power of diagnosis than average methylation.Moreover,the DNA methylation marker had significant differential prognostic risk in breast cancers with the same subtype.It is suggested that DNA methylation markers we identified may associated with heterogeneity of breast cancer,and can be highly accurately predict the prognosis of breast cancer risk.Hypermethylation in local regions such as Cp G island and promoter is one of the important epigenetic variation in cancer.However,DNA methylation patterns in the region had high diversity that may represent differential epigenetic regulation mechanism.We focused on the concordant methylation pattern of adjacent Cp Gs in the region and developed the algorithm(Cell Methy)based on sliding window and integra strategy.The concordant methylated regions can be identified and quantified in a single tissue.Cell Methy showed high power in identification and quantification of concordant methylated region using simulation data.We found that cancer genome had more,longer and higher level of concordant methylation region than normal genome through 11 normal and 12 cancer tissues.Compared to average methylation level,concordant methylation pattern reflect more divergence among tissue,especially specific in caner.Finally,we applied Cell Methy to two cancers including breast cancer and chronic lymphocytic leukemia,and found more regions showed significant differences of concordant methylation rather than average methylation,especially hyper-concordant methylation level in cancer.Even the region had lower average methylation in cancer that tend to show high concordant methylation level.These hyper concordant methylation region significantly enriched in the cancer related functions and pathways including regulation of molecular,cell death,reponse of cellular to stress and MAPK signal pathway.In addition,the differential concordant methylation regions were shared in multiple chronic lymphocytic leukemia indivudals,suggesting concordant methylation pattern may be the characristic in cancer.In this paper,we used the bioinformatics algorithms and statistical models to identify the DNA methylation markers associated with human phenotype in different scales from species to indivuals to tissues.These DNA methylation markers showed more differences between human and other species,inter-individuals and inter-tissues especially between cancer and normal.Moreover,these DNA methylation markers involved in functional regulation of phenotype-specific in human,which provide a new perspective and candidate markers.
Keywords/Search Tags:DNA methylation, Fixed effect, Directional selection, Canonical correlation, Concordant methylation pattern
PDF Full Text Request
Related items