Abstract¡ªA three-dimensional pharmacophore model was developed based on 22 currently available inhibitors, which were carefully
selected with great diversity in both molecular structure and bioactivity, for discovering new potent neuraminidase (NA) inhibitors
to fight against avian influenza virus. The best hypothesis (Hypo1), consisting of five features, namely, one positive ionizable
group, one negative ionizable group, one hydrophobic point, and two hydrogen-bond donors, has a correlation coefficient of 0.902,
a root mean square deviation of 1.392, and a cost difference of 72.88, suggesting that a highly predictive pharmacophore model was
successfully obtained. The application of the model shows great success in predicting the activities of 88 known NA inhibitors in our
test set with a correlation coefficient of 0.818 with a cross-validation of 98% confidence level. Accordingly, our model should be
reliable in identifying structurally diverse compounds with desired biological activity.