On the paradigm shift from rational to random design

Abstract

Lead identification and optimization are key elements of the drug discovery process, in which “rational molecular design” plays an important role. Its ultimate goal, the successful de novo design of high-affinity ligands based on three-dimensional structure information, is still only partially reached due to our limited understanding of nonbonding interactions and solvation effects. A plethora of highly sophisticated molecular modelling, biocomputing and graphics methods are available, providing powerful interactive environments for the semiquantitative molecular design. This structure-based approach is complemented by efforts of random screening of large compound repositories without preconceived notions on structure-activity relationships. This dual approach of “rational design and random screening” has established itself as an essential element in the overall drug discovery process. An interesting shift of paradigm has taken place towards “random design and rational screening”. Random design entails diverse methods of combinatorial modelling, fuzzy or sketchy design, and nonstructural design concepts. Screening efforts are no longer performed on extensive compound repositories comprising hundreds of thousands of compounds, but rather on biased subsets, rationally selected by substructural or pharmacophore hypothesis-based database searches. These efforts are complemented by new methodologies of macro- and microsynthetic combinatorial chemistry approaches with novel tools for structure diversity planning and assessment.