Protein Flexibility

A strategy that docks ligands against a discrete set of receptor conformations rather than a single static structure. These conformations can be sourced from multiple crystal structures, NMR ensembles, or molecular dynamics trajectories. By sampling diverse binding site shapes, ensemble docking captures pocket adaptability that rigid docking misses, significantly improving hit rates for targets like kinases and GPCRs where side-chain rotamer states gate ligand access.

A docking protocol that permits active site residue side-chain flexibility during the ligand placement process. Unlike rigid docking, the receptor is allowed to adjust its binding pocket geometry in response to the ligand's steric and electrostatic features. This is critical for targets where the apo structure occludes the true binding cavity. Induced-fit models prevent the dismissal of ligands that require conformational selection to achieve a high-affinity complex.

The computational detection of transient binding cavities that are absent in experimentally determined static structures but emerge due to thermal fluctuations. Methods like mixed-solvent molecular dynamics or Markov state models reveal these hidden sites. Targeting cryptic pockets can yield highly selective allosteric inhibitors for previously 'undruggable' targets such as KRAS G12C, where the switch-II pocket is only visible in dynamic simulations.

A rapid computational method for predicting the large-scale, low-frequency collective motions of a protein. By approximating the energy landscape as a harmonic potential, NMA identifies hinge-bending and domain-swinging movements that define functional transitions. It is computationally inexpensive compared to full MD and is often used to generate biologically relevant conformers for ensemble docking or to interpret cryo-EM heterogeneity.

Two fundamental models of ligand binding. Conformational selection posits that the unbound protein pre-exists in an ensemble of states, and the ligand selectively binds to a compatible conformation. Induced fit asserts that initial ligand contact triggers a structural rearrangement. In reality, most systems operate via a hybrid mechanism. Understanding which model dominates informs whether to screen against multiple pre-generated conformers or employ flexible docking algorithms.

The B-factor (or temperature factor) from X-ray crystallography quantifies the uncertainty in an atom's position due to thermal motion and static disorder. High B-factors indicate flexible loops or termini. In virtual screening, B-factor data can be used to weight docking grids, softening the repulsive van der Waals potential in highly mobile regions to prevent steric clashes from penalizing valid ligands that would otherwise bind to a dynamic loop.

A strategy that docks ligands against a pre-computed collection of discrete receptor conformations rather than a single rigid structure. These conformations may be sourced from multiple crystal structures, NMR ensembles, or molecular dynamics trajectories. By sampling diverse binding-site geometries, ensemble docking accounts for conformational heterogeneity and can identify ligands that bind to transiently accessible states. Key considerations include ensemble selection—how many and which structures to include—and consensus scoring across the ensemble to rank compounds.

A docking protocol that permits limited receptor flexibility by allowing active-site side chains—and occasionally backbone segments—to adjust in response to ligand binding. Unlike rigid docking, induced-fit methods iteratively sample residue conformations and ligand poses to find a mutually compatible complex. This approach is critical for targets where the binding pocket undergoes significant rearrangement upon ligand binding, such as kinases with a DFG-in to DFG-out transition. The computational cost is substantially higher than rigid docking but lower than full molecular dynamics.

Transient binding cavities that are absent in experimentally determined static structures but emerge due to protein thermal motion. These hidden sites are invisible to conventional rigid docking and represent high-value targets for allosteric modulation and targeting proteins previously considered undruggable. Identification relies on molecular dynamics simulations, enhanced sampling techniques, or specialized algorithms like FTMap and MixMD that detect surface concavities and favorable binding hotspots in dynamic trajectories.

A computationally efficient method for modeling large-scale, low-frequency collective motions of proteins. By approximating the energy landscape as a harmonic potential, normal mode analysis calculates the dominant vibrational modes that describe domain movements, hinge-bending, and channel opening. The elastic network model variant requires only C-alpha coordinates, making it applicable to large macromolecular assemblies. These modes are often used to generate biologically relevant conformations for ensemble docking without the expense of full molecular dynamics.

A binding mechanism in which a ligand selectively binds to a pre-existing, low-population conformation of the receptor, shifting the equilibrium toward that state. This contrasts with induced-fit, where the ligand actively reshapes the binding site. Conformational selection is supported by NMR relaxation dispersion and single-molecule FRET experiments revealing that proteins spontaneously sample their bound conformations in the absence of ligand. Virtual screening strategies must therefore sample these rare but functionally critical states to avoid false negatives.

A physics-based computational method that numerically solves Newton's equations of motion for all atoms in a solvated protein system over time. MD provides a full atomistic description of protein flexibility, capturing side-chain rotations, loop movements, and domain rearrangements at femtosecond resolution. For virtual screening, MD trajectories are clustered to extract representative conformational snapshots for ensemble docking. Enhanced sampling variants like replica exchange and metadynamics accelerate exploration of rare events and cryptic pocket opening.