A neural network to predict multiple functions of genes

A neural network to predict multiple functions of genes

Prediction of Multiple Gene Functions by Using Neural Networks Jiangsu Qian Chen Chunxian

Chang Gung Bioinformatics Center, Chang Gung University, Institute of Information Management

[email protected] [email protected]

Summary

With the human genome decoding, for health and medical attitudes in significant changes in symptoms to diagnosis and treatment of diseases, traditional medical methods, and with the growing

understanding of the genome, and gradually the genetic pathology, gene diagnosis and gene therapy mainly of "genomic medicine"

replaced. Then genomic era one of the important tasks is to

understand the genome, so the understanding of gene function for the post-genomic era one of the major issues.

As the complexity of the human genome and the lack of Wan

Zheng De Jiao related experiments, and Shiyan Shi Zhong Yao yeast as the model organism and sequence since 1996 after the

completion of Ding has abundant information on scale experiments,

this study is to yeast Duixiang, to the known structure of yeast gene

functions associated with the yeast gene expression data for yeast

gene function prediction, and for the use of cluster analysis to

analyze gene expression data can predict the weakness of a single

gene function, to back-propagation neural network classifier as the

basis for the development of a genetic analysis to be performed

multiple functions approach to living organisms closer to the actual

operation of the case, and to open the yeast genes with known

functions relevant information for training and learning genetic

information.

Keywords: gene function prediction, multiple gene function

prediction, classification, propagation neural network