Font Size: a A A

Protein Interaction Networks Experimental Comparison Analysis

Posted on:2012-10-26Degree:MasterType:Thesis
Country:ChinaCandidate:S WuFull Text:PDF
GTID:2210330368498350Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the completion of human genome sequencing, the research on protein-protein interaction has become one of very important issues. Using PPI information and predicting protein function is an important aspect.Its research has become a major field of life sciences. In this paper,we researched the methods of predicting the proteins function using PPIs and PIN.This article through the use of protein interactions network based on the forecasting method using bioinformatics, construct the function of protein interactions, using the network of protein interactions network new clustering method, establishes the cell function related model, clarify cell function molecular mechanism, predict protein function and the interaction, etc, in order to solve the problem of biology problem.For the more accurately predict protein function, this paper arithmetic average minimum value K - means clustering algorithm was improved, the improved AAMV means the K - the clustering algorithm function prediction of proteins. First of all, according to the protein-protein interactions, through human AD (Alzheimer Scotland warned against 's associated with) that PPI network diagram, the connection between matrix protein; Then use AAMV method for similarity matrix; Finally, on the basis of the similarity matrix is put forward in this paper, using the weighted error squares standards by the improved effectively, the convergence means clustering methods K - the proteins of PPI network clustering and function prediction.This paper designs experimental comparison of three kinds of the improved algorithm AAMV law means the K - Bridge and Korbel algorithm, Maryland. Analysis time complexity of three algorithm based on improved ABAAMV found the K - Maryland means algorithm and the complexity of the Bridge method are O (n), time complexity Korbel method for O (n4 interchange), experimental results found AAMV method based on improved means algorithm and the K - Maryland Bridge method to the inosculation results for 100%, while Korbel method with the inosculation results is 97.5%. So the ABAAMV based on improved means the K - the new algorithm has better time complexity and coincide degrees.
Keywords/Search Tags:Protein-protein interaction, The protein-protein interaction network, The functional prediction, The Clustering Algorithm
PDF Full Text Request
Related items