This book investigates the evolving landscape of quantitative structure-activity relationships (QSAR) and their application in drug design and chemical risk assessment. The text explores new molecular descriptors derived from three-dimensional structures, integrating classical regression and PLS analysis with modern nonlinear techniques like neural networks and support vector machines. It aims to delineate the scope and limitations of various modeling approaches. Case studies from pharmacology, toxicology, and ecotoxicology are utilized to illustrate the practical utility of each method.

The book addresses the rapid growth and evolution of QSAR techniques over the past 15 years, highlighting their increasing importance in drug design and chemical risk assessment. It focuses on the integration of 3D structural information with both traditional and advanced computational methods. The inclusion of case studies from pharmacology, toxicology, and ecotoxicology suggests a practical and applied approach to understanding these complex modeling techniques.