The concept of chemometric guidelines in synthesis and development of new anticancer compounds

Abstract

© 2017 by Nova Science Publishers, Inc. All rights reserved. The discovery and development of new anticancer drugs are quite complex and timeconsuming tasks. In past decades researchers have been trying to minimize the number of trial-and-error experiments in certain stages of drug development by analyzing and interpreting already obtained experimental results by using chemometrics. It has been shown that chemometrics is a very useful tool in the early stages of drug development. Quantitative Structure-Activity Relationship (QSAR) analysis, as one of the main chemometric approaches, is based on the correlation between molecular descriptors and biological activity. QSAR analysis can point out which molecular characteristic(s) can be crucial for the final biological response. Therefore, the QSAR models can be considered to be valuable guidelines for the synthesis of new compounds which will express the desired biological activity. The present chapter describes the main steps in the formation of QSAR guidelines for the synthesis of novel anticancer compounds based on androstane structure, as well as a new approach in the ranking and selection of the best QSAR models. The significance of the multivariate chemometric approaches is particularly emphasized, including multiple linear regression, principal component regression, partial least squares and artificial neural networks. The significance of pattern recognition methods, such as principal component analysis, in drug discovery process is described as well. This chapter is mostly focused on the application of chemometrics in the development of androstane-based anticancer compounds and the prediction of anticancer activity of newly synthesized androstane derivatives.