A Novel Modeling Protocol for Protein Receptors Guided by Bound-Ligand Conformation

Abstract

A novel protocol for protein homology modeling is described, in which uncertainty in protein structure is resolved by applying the criterion that the protein must bind to a complementary ligand. A diverse library of protein models is created and then screened by docking with ligands of known conformation. The more accurate protein models form higher-quality docked complexes, and the quality of the fit is used to select the best models. The effectiveness of this technique with both natural and unnatural ligands is demonstrated by modeling the Fv fragment of an antibody and comparing the results to known crystal structures.

Discovery of Novel Chemotypes to a G-Protein-Coupled Receptor through Ligand-Steered Homology Modeling and Structure-Based Virtual Screening

“…Because of the lack of accurate binding energy calculations, limitations involved in the modeling of the system (e.g., lack of phospholipids and water molecules, lack of abundant mutagenesis data, absence of a wealth of solid experimental evidence to be used as structural constraints), and lack of and uncertainties in ligand-protein interaction patterns in many systems, especially in GPCRs (cf. ref 50 where models are selected according to docking score and quality of binding pose), we felt it was better to validate the final set of models through the performance of the modeled binding sites in a smallscale virtual screen. In this way, we test models in the demanding experiment of discriminating between known MCH-R1 antagonists described previously 27 and GPCR class A binders.…”

“…The nomenclature of Ballesteros and Weinstein 67 is used, whereby the most conserved residue in helix X is labeled as X. 50.…”

Ligand-supported Homology Modelling of Protein Binding-sites using Knowledge-based Potentials

“…The more relevant protein models achieved better-scored docking solutions, and the quality of the binding modes generated was thus used to select the most relevant modes. 40 This approach has been applied to the modelling of F v antibody fragments. However, it requires experimental data about the conformation of the docked ligand.…”

Relative free energy of binding and binding mode calculations of HIV‐1 RT inhibitors based on dock‐MM‐PB/GS

“…An alternative way to predict the binding structure of a substrate in its receptor is by a docking simulation, which has been successfully used in many applications 24–32. Some docking methods have demonstrated promising power to predict reasonable binding structures.…”