Improving Activity Prediction of Adenosine A2B Receptor Antagonists by Nonlinear Models

摘要

This study deals on estimation of ligand activity with its descriptors. So, to achieve this goal, two different approaches were implemented. In the first one, the intervals between samples were determined. But in the second method, the intervals were clustered with k-means method. Afterwards, best descriptors of each ligands were extracted with genetic algorithm. Then, observations were classified with One-Against-All method. Finally, the activity of each ligands were estimated by forty percent of samples. In the first method, AUC values were between fifty four to ninety seven percent. For second approaches, there were about ninety seven percent.