Abstract

The analysis of the influence of molecular features which can be extracted from the simplified molecular input line entry system (SMILES) and involved in the process of the building up of a series of QSAR models (with different splits into training and test sets) by means of the CORAL software for mutagenicity and anticancer activity has been performed. The presence of nitrogen (sp3) is favorable for decrease of the both endpoints; the presence of only one ring is also promotor for decrease of the both endpoints; however the presence of two or three rings is favorable for increase of mutagenicity and decrease of anticancer activity. These findings provide useful criteria for further experimental and computational studies in the search for new anticancer agents.

Keywords: Anticancer activity, Mutagenicity, QSAR, SMILES, Validation, CORAL software, polymer film, anticancer agents, molecular, Monte Carlo optimization.