DockingVersus Pharmacophore Model Generation: A Comparison of High-Throughput Virtual Screening Strategies for the Search of Human Rhinovirus Coat Protein Inhibitors
Theodora M. Steindl, Thierry Langer · QSAR & Combinatorial Science · 2005
Inhibition of the human rhinovirus coat protein is an intensively studied approach towards finding a therapy for common cold. We present the results of a docking study using the tool LigandFit and a database of 1000 drug-like compounds seeded with known active HRV coat protein inhibitors. One scoring function performed especially well in ranking the active molecules at high positions. The conformations and positions from the docking process were found to be in good agreement with the X-ray data. Furthermore, highly selective pharmacophore hypotheses were generated in a structure-based approach. They include information on necessary chemical functions as well as on the required shape. When applied in 3D database screening, they selectively retrieved compounds with known antirhinoviral activity from large data sets. We compare these two virtual screening methods and summarise their benefits but also the observed disadvantages. Finally, a focussed virtual library for a special class of rhinovirus coat protein inhibitors was generated with our novel tool ilib diverse and screened with the pharmacophore models described.