Association Studies on mtDNA and Parkinson’s Disease Population Discrimination Using the Statistical Classification
Jun Wang, Le Zhang, Quan Zou, Jun Tan, Xiaowei Chen and Yukun WuAffiliation:
Department of Biostatistics and Computational Biology, University of Rochester, 601 Elmwood Avenue Box 630, Rochester, New York 14642, USA.
AbstractSince mitochondrial DNA (mtDNA) follows directly maternal inheritance, the presence of common polymorphisms in mtDNA sequences can classify mtDNAs into haplo groups and sub-haplo groups. Regarding to the rapidly growth of mtDNA sequences, many bioinformatics scientists are dedicated to uncover the association between the common mtDNA polymorphisms and the complex genetic diseases. In this study we analyze the mtDNA sequences from 96 Japanese Parkinson’s disease (PD) patients and 96 Japanese normal persons. A special algorithm based on keyword tree is employed to quickly align the mtDNAs. The mitochondrial single nucleotide polymorphisms (mtSNP) are revealed from the mtDNA alignments by using the genetic characteristic of SNPs and mtDNAs. A statistical significance based locating algorithm is proposed to select the disease associated mtSNPs as the features of classification in disease association research. Sequence transforming probability is introduced in the process of sample classification to discriminate the Parkinson’s disease patients and the common persons. The experimental results indicate that Parkinson’s disease patients can be characterized by unique mtSNPs. Although several mtSNPs are different from previously reported mtSNPs, the algorithm precision of Parkinson’s disease population discrimination reaches 90%.
Classification, genetic association analysis, mitochondrial DNA, mitochondrial SNP, Parkinson’s disease, statistical significance.
Purchase Online Rights and Permissions