Wavelet and PCA for Metabolic Network Comparison

This paper is composed based on principal component analysis and wavelet transform method to compare the metabolic networks. We combine 7 centrality topologies of metabolic network to extract a comprehensive center on principal component analysis and construct the original signal based on statistical method. By using the wavelet tool to describe the metabolic network structure, we build the fuzzy function to compare the metabolic network structure between different species. This method above can effectively analyze the structure similarity of different species´ metabolic networks, reveal the species-specific of metabolic networks and lay a mathematical foundation for the evolution of metabolic networks. Also it is the first time to use the wavelet analysis method in studying the metabolic networks, and it makes an attempt on a new research direction.