Modeling and Analysis of Nucleosome Positioning Signals

This project aims to develop signal processing based methods for the detection and analysis of nucleosome positions along a DNA sequence. The research team proposes the matched mirror position transform (MMPT) to model and extract weak spectral components from the positioning signals. The researchers will investigate a relaxation based technique to remove impulse noise in the observed signals and enhance the performance of the MMPT. The researchers will study different types of sequence features, parametric models and multi-scale approaches for spectral analysis of the positioning signals. The researchers will also develop robust methods for data classification. Their methods will be tested with large genomic datasets.Nucleosomes can play a very important role in gene regulation and the investigation of their properties and functions has attracted enormous interests from many researchers. It is costly to conduct wet-lab experiments to determine nucleosome positions and computer techniques can provide an effective solution to many related research problems. Their work can lead to a better understanding of nucleosome positioning signals and scientific and technical advances in modeling and analysis of such signals. In addition, this research can result in useful computer tools for nucleosome position prediction and find many other practical applications to genomic data analysis in general.