Topological models for prediction of physico-chemical, pharmacokinetic and toxicological properties of antihistaminic drugs using decision tree and moving average analysis.

Various topostructural and topochemical indices were used to encode the structureal features of antihistaminic drugs. The values of 18 indices for each drug comprising the dataset were computed using an in-house computer program. In the present study, decision tree and moving average analysis were used to predict physico-chemical (log P), pharmacokinetic (T(max)) and toxicological properties (LD(50)) of antihistaminic drugs. A decision tree was constructed for each property to determine the importance of Topological Indices (TIs). Single topological index based models were developed using moving average analysis. The tree learned the information from the input data with an accuracy of >94% and predicted the cross-validated (10-fold) data with an accuracy of upto 71%. Moving average analysis resulted in single index based models with an accuracy upto 80%.