Incorporating Protein Flexibility in Virtual Screening

GATICA, EDGAR A.; CAVASOTTO, CLAUDIO N.

Orlando, FL

Conferencia; Annual Biomedical Research Conference for Minority Students; 2008

Typical structure-based virtual screening uses a flexible ligand-rigid receptor docking approach.  This method fails to account for the induced-fit effects, in which the receptor flexes upon ligand binding.  As a result, the docking algorithm may fail to dock a known binder to a rigid receptor.  Incorporating receptor flexibility may be necessary to ensure proper docking in cases where rigid receptor docking is unsuccessful.  However, explicit flexible-ligand:flexible-receptor docking is computationally too expensive for use in large-scale virtual screening experiments.  An alternative strategy, which has shown impressive results, is the use of multiple rigid receptors to simulate a single flexible one.  The docking scores are then merged, and the resulting list is condensed to a size equivalent to that of a single screening (the merging-shrinking procedure).  Unfortunately, previous attempts to use this method have shown mixed results.  We hypothesize that this is due to the inclusion of redundant receptor combinations, that is, combinations of too similar receptors, what does not improve –but dilutes- the performance of the method.