The Dynamic Epigenome: Analysis of the Distribution of Histone Modifications

There is a genome in a cell, as everyone knows, but there is also an epigenome. The epigenome regulates the transcription of the underlying genome. In the last decade, it was discovered that the epigenome state and its regulation are important for differentiation and development. Correlation studies with aging samples had led to the hypothesis that misregulation of the epigenome causes aging and cancer. Furthermore, diseases were identified which are caused by errors in the epigenome state and its regulation. Identification of erroneous epigenome states and misregulation requires the prior knowledge of the common state. Several studies aim at measuring epigenome states in different organisms and cell types and thus, provide a huge amount of data. In this dissertation, a pipeline is developed to analyze and characterize histone modifications with respect to different cell types. Application of this pipeline is shown for a published data set of mouse consisting of data for H3K4me3, H3K27me3, and H3K9me3 measured in embryonic stem cells, embryonic fibroblasts and neuronal progenitors. Furthermore, methods for the detection of the epigenetic patterns are presented in this dissertation. Therefore, a segmentation method is developed to segment the genome guided by the data sets. Based on this segmentation, the epigenome states as well as epigenetic variation can be studied. Different visualization methods are developed to highlight the epigenetic patterns in the segmentation data. Application of the segmentation AND visualization methods to the mouse data set had resulted in not only colorful squares but also in biological conclusions! It demonstrate the power of the developed methods. Although the studied data set in this dissertation contains only ordinary tissue cells, the methods are not restricted to study the reference epigenome state. Comparison of normal and disease cells as well as comparison with aged cells are possible with all of the methods. Finally, the methods are compared based on the obtained results. It shows that all methods highlight different aspects of the data. Thus, applying all methods to the same data sets, deep insights into the epigenome in murine embryonic stem cells, embryonic fibroblasts and neuronal progenitor cells are gained. For example, it had been found that several mechanisms exist setting H3K4me3 marks. Furthermore, not all mechanisms are found in all cell types. Strong evidence had been found that catalysis of H3K4me3 and H3K27me3 is coupled