Protein Function Prediction Based On Clustering Algorithm On A Protein-protein Interaction Network

Protein-protein interaction network refers a new research area of computer science. Prediction of protein function is a challenging problem at present in the research of protein-protein interaction network. Any new breakthrough in the research will be conducive to expounding the chang mechanism of organism under physiological condition directly. What is more, it will be a significant assistant to relevant industries such as biomedical engineering, agbio-tech, etc.On the basis of thorough analysis of existing clustering algorithm in protein-protein interaction network and protein function prediction method, we develop our work on the study of clustering algorithm suited protein-protein interaction network and protein function prediction. We propose a novel clustering algorithm and perform series of experiments. We associate ant colony algorithm with our clustering result to predict the protein function. We achieve good results. The main contributions of this thesis are summarized as follows:We take a closer look at the difficulties and challenges that protein function prediction is facing now and propose the research scheme of this thesis. We survey in the existing protein interaction, properties and characteristics of PPI network. A thorough analysis of protein function prediction from protein sequence, structure and protein-protein interaction is given.The distance metric in such setting was redefined by the network distance,in the past, which has to be computed by the expensive shortest path distance over the network.The existing methods are not applicable to such cases.Therefore,by exploiting unique features of networks,a new clustering algorithm is presented,which uses the information of nodes and edges in the network to prune the search space and avoid some unnecessary distance computations.The experimental results indicate that the algorithm achieve high efficiency for clustering nodes in real protein-protein interaction network.Using ant colony algorithm, we make an alignment between the orphancy protein sequence and the protein sequence of clustering center to predict function of orphancy protein. Experimental results show that this method is reasonable and efficient. And compare our aligning results with the results of the aligning sliding window methods, we can get satisfying results.