ChIP-seq Data Plays an Important Role in a Cytosine-Based DNA Methylation Prediction Model

DNA methylation was found previously related with histone modifications. The relationship among DNA methylation and histone modifications is potentially identifiable by integrating ChIP-seq and methylation data. However, little has been addressed on this issue in literature. Predicting DNA methylation can largely avoid the high cost of bisulfite-converted DNA experiments and drawbacks of micro array-based approaches. The most important, biological studies can be designed in a more economical manner. In this study, we found the DNA methylation was strongly influenced by surrounding combinatorial enriched ChIP-seq derived features in genome-wide scale. As an application, a cytosine-based methylation prediction model is proposed to predict the methylation status. As a result, we found Lymph-specific genes were distinct from other kinds of genes around Transcription Start Sites, which confirmed that our model is tissue-specific.