Prediction of Potential Antimigraine Activity Using Artificial Neural Networks

Authors

Qing Luo, University of Arkansas at Little RockFollow
Jerry A. Darsey, University of Arkansas at Little RockFollow
Cesar M. Compadre, University of Arkansas at Little RockFollow
Sanjay K. Mitra, University of Arkansas at Little RockFollow

Abstract

More than 10 million Americans, three quarters of them women, suffer some degree of recurrent migraine headaches. Feverfew [Tanacetum parthenium (L.) Schultz-Bip.) is a member of the Asteraceae family that is native to Europe. This plant is a perennial flowering aromatic plant common in gardens. It has been widely used as a self-medication of arthritis, fever, and migraine headaches for over 2000 years. Sesquiterpene lactones (SL) are the components responsible for the antimigraine activity of feverfew. In this research, the relationship between SL structural information and their biological activity was studied by using Gaussian 92 program in conjunction with artificial neural networks (ANNs).The molecular orbital parameters of SL were obtained by using Gaussian 92 program. A set of 39 SL molecules was divided into two groups, a training set containing 33 molecules and a testing set containing six molecules. An ANN was trained and tested by using training sets and testing sets on SL's antimigraine activities. The results showed that ANNs successfully predicted the antimigraine activities of SL based on their different structural information.

Recommended Citation

Luo, Qing; Darsey, Jerry A.; Compadre, Cesar M.; and Mitra, Sanjay K. (1997) "Prediction of Potential Antimigraine Activity Using Artificial Neural Networks," Journal of the Arkansas Academy of Science: Vol. 51, Article 20.

Available at: https://scholarworks.uark.edu/jaas/vol51/iss1/20