Objective models for steroid binding sites of human globulins

Abstract

We report the application of a recently developed alignment-free 3D QSAR method [Crippen,G.M., J. Comput. Chem., 16 (1995) 486] to a benchmark-type problem. The test systeminvolves the binding of 31 steroid compounds to two kinds of human carrier protein. Themethod used not only allows for arbitrary binding modes, but also avoids the problems oftraditional least-squares techniques with regard to the implicit neglect of informative outlyingdata points. It is seen that models of considerable predictive power can be obtained even witha very vague binding site description. Underlining a systematic, but usually ignored, problemof the QSAR approach, there is not one unique type of model but, rather, an entire manifoldof distinctly different models that are all compatible with the experimental information. Fora given model, there is also a considerable variation in the found binding modes, illustratingthe problems that are inherent in the need for ’correct‘ molecular alignment in conventional3D QSAR methods.