Antibodies (Abs) play a vital role in many biological processes, most famously immunity and infection; often, their role is achieved by binding to pathogenic peptide sequences. Elucidation of Ab-peptide recognition elements is a foundation for the development of novel antigens and Abs, potentially contributing to immuno-suppressive or immuno-stimulatory treatment. While traditional structural investigations by X-ray crystallography and NMR offer a wealth of information, computational docking is a faster and more affordable structural technique, readily applicable to any target, and is complementary to experimental investigations. In this work, we present a validation study of several computational docking approaches to determine their suitability for studying Ab-peptide systems. Two docking programs, Glide and GOLD, were tested, and variations in scoring schemes and conformational sampling were investigated. It was found that additional conformational sampling greatly improved docking accuracy, but only for systems involving peptide ligands longer than 6 residues. Recent improvements to Glide’s docking protocol include a polypeptide-specific methodology; validation of this methodology may reveal greater capabilities in Ab-peptide docking.

A comparison between docked ligands and crystal structure ligand showing the effects of additional conformational sampling. (a) Crystal structure (green), Glide Standard Precision (blue), Glide Extra Precision (red). (b) Crystal structure (green), Glide Standard Precision + MacroModel (yellow). System is SYA/J6 Fab fragment complexed with an octapeptide O-antigen mimic (PDB: 1PZ5).