EFFICIENT MOLECULE REDUCTION FOR DRUG DESIGN BY INTELLIGENT SEARCH METHODS

Abstract

Search methods applied to data mining techniques help us to analyze a data set. These methods are used to reduce the size of the search space in order to select the relevant molecules for the drug design. The research community in theoretical chemistry is very much depends on practical prediction and classification tools for this purpose. Classification is one of the major data mining methodologies. The objective of this paper is to check the learning algorithms for classification of drug design parameters based on ‘Musk’ qualifying dataset. The main intention in this context is to deal with a large data set with high accuracy. For this purpose Bayesnet, Naïve-Bayes, Decision table and random forest models are built using Weka tool under supervised learning algorithm.  It is necessary to reduce the data dimension before constructing the models and thus the search methods for selection of attributes are followed. Those models are to be applied to predict the possible new drug candidates.