Extron prediction method based on improved period-3 feature strategy

Abstract

To improve the accuracy of the gene encoding (exon) prediction, near period-3 feature exons prediction algorithm is proposed. Near period-3 clustering power spectrum of extrons and introns are extracted as template feature, DNA sequence is divided into frames and moved. Compared with the template feature, the prediction of the Euclidean distance with different weights is realized from each frame. By changing the different feature, number, frame length, gene sequence weight and comparing with period-3 algorithm, the experiment results show that the prediction accuracy of the proposed algorithm is better than that period-3 algorithm.