Drug Design - Assessment of Molecular Similarity and Diversity


Major goal of combinatorial synthesis is to obtain compounds that have a wide structural variety while simultaneously covering the structural and chemical space as completely as possible. Thereby, the probability to find active compounds within the newly created chemical libraries is increased. On the other hand biological activity is intimately tied to molecular surface properties such as the molecular electrostatic potential. These surface properties can be encoded by spatial autocorrelation functions and projected into two-dimensional maps with Kohonen neural networks. With the Kohonen maps the structural and chemical space, described by the libraries under consideration can be visualized. This visualizition allows an estimation of the similarity and diversity of the libraries. Based on that estimation, recommendations for a proper selection of libraries can be given.


M. Wagener, J. Sadowski, J. Gasteiger,
Autocorrelation of Molecular Surface Properties for Modelling Corticosteroid Binding Globulin and Cytosolic Ah Receptor Activity by Neural Networks,
J. Am. Chem. Soc., 1995, 117, 7769-7775.

J. Sadowski, M. Wagener, J. Gasteiger,
Assessing Similarity and Diversity of Combinatorial Libraries by Spatial Autocorrelation Functions and Neural Networks,
Angew. Chem. Int. Ed. Engl., 1995, 34, 2674-2677.


Dr. Lothar Terfloth