A dynamic programming algorithm for nucleosome positions alignment

Abstract

Nucleosomes are the basic units of eukaryotic chromatin. The nucleosome positioning is dynamic for various cell types and biological states, resulting in specific gene regulation. Currently, there is no approach to find the correspondence between two sets of nucleosomes to reveal the difference of their positions. We develop a method for nucleosome positions alignment based on the dynamic programming algorithm, which can quantify the changes in nucleosome locations with scores and evaluate regional dynamics changes including translation and missing. Given the result of a peak list stands for nucleosome positions, to align the peaks from two samples, our method accumulate all pair scores for match, replacement or deletion and choose the maximum one as the optimal alignment. From nucleosome alignment we can find one-by-one correspondence between nucleosome positions in different cell stages and the conservative stable and variable regions, which can be used to recognize dynamic behaviors of nucleosome shift and eviction.