β-secretase (BACE1) inhibition has emerged as a most promising target for the treatment of Alzheimer’s disease.
In the present study an in silico approach has been successfully utilized for the development of diverse classification
models for the prediction of BACE1 inhibitory activity using a dataset consisting of 42 differently substituted aminohydantoin
analogues. Classification tree (CT), moving average analysis (MAA) and random forest (RF) were utilized for development
of models. Two out of three MDs identified by CT as the most important were the detour cum distance matrix
based topological descriptors proposed in part-I of the manuscript.
Various models resulted in the prediction of BACE1 inhibitory activity with an overall accuracy of >92%. Overall accuracy,
non-error rate, intercorrelation analysis, specificity, sensitivity and Mathew’s correlation coefficient (MCC) were
utilized to determine statistical significance of the said models. Proposed models provide an immense potential for furnishing
lead molecules so as to unfold potent BACE1 inhibitors for the treatment of Alzheimer’s disease.