Journal of Applied Sciences ›› 1997, Vol. 15 ›› Issue (2): 217-222.

• Articles • Previous Articles     Next Articles

SELF-ORGANIZATION NEURAL TREE APPLIED TO STRUCTURE-ACTIVITY RELATIONSHIP OF FENTANYL DERIVATIVES

LU WENCONG1, FANG JIANHUI1, DING YIMING1, YAN LICHENG1, CHEN NIANYI2   

  1. 1. Shanghai University, Jiading Compus, Shanghai 201800;
    2. Shanghai Institute of Metallurgy, Chinese Academy of Sciences, Shanghai 200050
  • Received:1995-06-30 Revised:1996-05-11 Online:1997-06-30 Published:1997-06-30

Abstract: In this paper, the non-linear relationship between the characteristic parameters and analgesic activities of a series of fentanyl derivatives is calculated by self-organization neural tree model. The results show that the performance of the self-organization neural tree is good and the successful claasification rate is rather nigh. Therefore, the model might be used ac an effective assistant technique for the investigation of structure-activity relationship of drugs.

Key words: fentanyl derivatives, structure-activity relationship, artificial neural network, self-organization neural tree