Journal of Applied Sciences

• Articles • Previous Articles     Next Articles

Prediction of Transmembrane Helical Segments in Membrane Proteins Based on Discrete Wavelet Transform

YU Bin1 , WANG Guang-bin1, WANG Yi-fei2   

  1. 1. College of Mathematics and Physics, Qingdao University of Science and Technology, Qingdao 266061, China;

    2. College of Science, Shanghai University, Shanghai 200444, China

  • Received:2007-05-15 Revised:2007-08-08 Online:2008-01-31 Published:2008-01-31

Abstract: Increasing protein sequences from the genome project require theoretical methods to predict transmembrane helical segments (TMHs). In this paper, a method based on discrete wavelet transform (DWT) is developed to predict the number and location of TMHs in membrane proteins. PDB coded as 1KQG is chosen as an example to describe the prediction process with this method. Eighty proteins with known 3D-structure from the Mptopo database are chosen at random as the test objects (including 325 TMHs), 312 of which can be predicted and the prediction accuracy is 96.6%. Compared with main prediction results of seven common prediction methods, the obtained results indicate that the proposed method has higher prediction accuracy.

Key words: discrete wavelet transform, membrane protein, transmembrane helical segments, hydrophobicity