The Application of Fisher Linear Discriminant to Distance Between Genomes Which Based on COGs
DOI:
Author:
Affiliation:

Clc Number:

Fund Project:

This work was supported by grants from The National Natural Sciences Foundation of China (39890070,19890380,39993420), Knowledge Innovation Project of The Chinese Academy of Sciences (KSCX2-2-07 and KSCX1-08) and a Special Grant Science and Technology Com

  • Article
  • |
  • Figures
  • |
  • Metrics
  • |
  • Reference
  • |
  • Related
  • |
  • Cited by
  • |
  • Materials
  • |
  • Comments
    Abstract:

    A new method to construct a phylogeny tree based on whole genome information is introduced. Each gene of an organism is represented by a 17 dimensional vector, each dimension of which relates to one of the 17 COGs(clusters of orthologous groups of proteins) classes. All the vectors of a genome constitute a set. Then Fisher linear discriminant was used to find a set of optimal weights which reflect more accurately the different contribution of the 17 COGs classes to the genome's evolution. That is, under the Fisher criteria, each vector of a genome is linear mapped. After that, the distance between two genomes was represented by the distance between the related two sets constituted by mapped vectors. At last, the distance matrix was used to construct a phylogenetic tree by PHILP software package. Phylogeny trees of 38 and 43 genomes constructed by this method respectively well support the “three primary kingdom” theory of Woese. This method rectifies the shortcoming of other methods which are difficult to compare genomes differring remarkably in genome size. In addition, the method diminishes the distortion on the distances between genomes brought by lateral gene transfer.

    Reference
    Related
    Cited by
Get Citation

LIU Rong, WANG Yue-Lan, ZHU Xiao-Peng, LING Lun-Jiang, HAN Ru-Shan. The Application of Fisher Linear Discriminant to Distance Between Genomes Which Based on COGs[J]. Progress in Biochemistry and Biophysics,2002,29(5):760-765

Copy
Share
Article Metrics
  • Abstract:
  • PDF:
  • HTML:
  • Cited by:
History
  • Received:March 06,2002
  • Revised:April 11,2002
  • Accepted:
  • Online:
  • Published: