文章基本信息

标题：Minimum Spanning Tree-based Clustering Applied to Protein Sequences in Early Cancer Diagnosis
本地全文：下载
作者：Dr. T. Karthikeyan ; S. John Peter ; B. Praburaj 等
期刊名称：International Journal of Computer Science & Technology
印刷版ISSN：2229-4333
电子版ISSN：0976-8491
出版年度：2012
卷号：3
期号：1
页码：642-648
语种：English
出版社：Ayushmaan Technologies
摘要：Cancer molecular pattern effcient discovery is essential in themolecular diagnostics. The number of amino acid sequence isincreasing very rapidly in the protein databases, but the structure ofonly some amino acid sequences are found in the protein data bank.Thus an important problem in genomics is automatically clusteringhomogeneous protein sequences when only sequence informationis available. The characteristics of the protein expression data arechallenging the traditional unsupervised classifcation algorithm.In this paper we use Minimum Spanning Tree based clusteringalgorithm for clustering amino acid sequences. A similarity graphis defned and a cluster in that graph corresponds to connected subgraph. Cluster analysis seeks grouping of amino acid sequence into subsets based on Euclidean distance between pairs of sequences.Our goal is to fnd disjoint subsets, called clusters, such that twocriteria are satisfed: homogeneity: sequences in the same clusterare highly similar to each other and separation: sequences in thedifferent clusters have low similarities to each other. A thoroughunderstanding of the genes is based on upon having adequateinformation about the proteins. Solving the protein relatedproblem has become one of the most important challenges inbioinformatics. In bioinformatics, number of protein sequencesis more than half million, and it is necessary to fnd meaningfulpartition of them in order to detect their functions. The methodwhich can enhance the structural recognition, classifcation andinterpretation of proteins will be advantageous. Many methodshave been adopted to solve such bioinformatics problem. OurMinimum Spanning Tree based clustering algorithm is useful andeffcient method in the collective study of protein subset. The keyfeature of the algorithm is ability to predict the 3D structure ofthe unknown protein sequence.