| Researches on proteomics has always occupied an important position in biological research,and research on protein complexes is also in-depth and has made breakthrough progress.Protein complex is formed by the combination of multiple proteins,which is the basis of exploring complex life processes.PPI network can be regarded as composed of multiple protein complexes.However,due to the noise of PPI data,the recognition accuracy is not high.Therefore,recognizing protein complexes and understanding their functional properties have become a hot issue in biology.This problem can be solved by constructing PPI weighted network and designing protein complex recognition algorithm.This thesis completes the following two works:(1)Aiming at the problems of PPI data quality and insufficient consideration of biological characteristics,this thesis proposes a weighted network construction method–TSSN based on topological structure characteristics and biological characteristics information.The similarity between proteins is calculated by the edge aggregation coefficient and the GO term information,and it is used as weights to construct PPI weighted network.Experiments show that this method filters the noise data and considers the biological characteristics,finally obtains better recognition effect.(2)Aiming at the problems of traditional algorithm mainly consider the topology of the network and ignore the structure of protein complex itself,a protein complex recognition algorithm based on neighbor node extended clustering–NNEC algorithm is proposed.The weights of the nodes are calculated and arranged in descending order to form a set of seed nodes,and the first-order neighbors of the seed nodes are traversed;then,the nodes are extended in the second-order neighbor graph;finally,according to the compactness of the composite,the sub modules are merged to complete the excavating of the complexes.Experiments show that the algorithm can identify more protein complexes more accurately. |
PDF Full Text Request