Peak detection of histone modifications based on ChIP-seq data in human genome

Histone modifications include histone methylation, acetylation, phosphorylation, and so on. Variable combinations of modifications types, the appearance time, and the position, are of great biological significances. These important epigenetic marks can be regarded as “Histone code”. Histone-DNA interactions can determine the cells´ gene expression level. Sequencing data measured by ChIP-Seq experiments carry great challenge for researchers in forecasting or positioning the true binding site called “Peak”. Peak detection is the basis of other further studies. In this paper we present a stable algorithm to detect the histone-DNA biding site. Using signal processing method for reference, we designed a five-step algorithm to achieve peak detection. Genome wide maps of 38 histone modifications were drawn by our algorithm. By changing the main parameters, we achieved the best results. Those modification maps could be used in analyzing the combinational pattern of modifications or the relationship between epigenetic marks and diseases.