MetaData

Application created by Morgan Nance with the original version created by Dr. Jason Labonte. See https://doi.org/10.1002/jcc.24679 for the details on the original glycan_dock application.

Please cite: Development and Evaluation of GlycanDock: A Protein-Glycoligand Docking Refinement Algorithm in Rosetta; DOI: 10.1021/acs.jpcb.1c00910

PIs: Dr. Jeffrey Gray (jgray [at] jhu [dot] edu)

Description

App: GlycanDock

GlycanDock is a high-resolution (i.e., full-atom only), Monte Carlo-plus-Minimization protein–glycoligand docking protocol. GlycanDock treats glycan chains (glycoligands) as flexible oligosaccharides (rather than as a small-molecule-type ligand with discrete "rotamers"). The GlycanDock algorithm begins with an input glycoligand conformation and randomly perturbs it in both rigid-body and glycosidic torsion angle space to promote conformational diversity in each independent docking trajectory. Next, a set of inner refinement cycles alternates between rigid-body and glycosidic torsion angle sampling followed by protein and sugar side chain optimization at the interface and full-complex energy minimization. To promote thorough sampling of local conformational space, the inner refinement cycles are wrapped in a set of outer cycles that ramp down the van der Waals attractive weight and ramp up the Lennard–Jones repulsive weight of the REF2015 scoring function. Thus, the initial search allows clashes and promotes diversification, while the late stages enforce rigid sterics—a strategy shown to be effective in FlexPepDock, Rosetta's protein–peptide docking algorithm.

GlycanDock can dock mono-, oligo-, and polysaccharides to protein receptors of interest. Glycoligands with branched or exocyclic glycosidic linkages and/or various chemical modifications (e.g., N-acetylation, methylation) are tolerated. GlycanDock using the the CHI (CarboHydrate Intrinsic) energy term to ensure that glycosidic dihedrals falls within known energetically-favorable torsion space. GlycanDock can refine a putative protein–glycoligand complex with relatively high confidence in the initial orientation of the glycoligand with respect to the binding site (via the -refine_only flag, which enables small perturbations during glycoligand docking refinement). GlycanDock can also be used to find both the orientation and the conformation of the glycoligand if only the relative protein binding site is known (default behavior). GlycanDock is not a global docking protocol, meaning GlycanDock only searches local conformational space around the starting position of the glycoligand.

See Working With Glycans for more information on carbohydrate modeling in Rosetta.

Algorithm

GlycanDock employs ten inner cycles of Monte Carlo sampling and optimization of the glycoligand conformation at the protein receptor interface. The inner refinement cycles are wrapped by ten outer cycles that ramp the weights of the attractive (fa_atr) and repulsive (fa_rep) terms in the Rosetta scoring function. Each of the ten inner cycles performs two sets of sampling and optimization procedures on the glycoligand: a set of eight rigid-body perturbations and a set of eight glycosidic torsion angle perturbations (performed in either order every cycle). Every perturbation is followed by interfacial side-chain rotamer optimization, and every other perturbation is followed by full-structure energy minimization.

Rigid-body sampling consists of uniform perturbations to the glycoligands center-of-mass as well as occasional translation of the glycoligand toward the protein receptor’s center-of-mass. This latter “sliding” step ensures that the glycoligand does not drift far away from the protein during the docking trajectory. Glycosidic-linkage sampling includes performing uniform and non-uniform perturbations of various magnitudes on randomly selected glycosidic torsion angles.

Full Description:

///@brief Main mover for GlycanDockProtocol using default behavior and settings
///
/// Outer cycles controlling score term ramping
///   Cycle 1: fa_atr 1.0 * 3.25  fa_rep 0.55 * 0.45455
///   ...
///.  Cycle 10: fa_atr 1.0  fa_rep 0.55 (default REF2015 values as of Feb 2021)
///
/// Inner cycles controlling glycoligand sampling and optimization
///   Cycle 1-10: Either
///                  8 rigid-body perturbations followed by 8 glycosidic torsion angle perturbations
///                or 8 glycosidic torsion angle perturbations followed by 8 rigid-body perturbations
///               After each single rigid-body or glycosidic torsion angle perturbation
///                pack side-chain rotamers
///                  RotamerTrialsMover at interface for rounds 1–7
///                  RotamerTrialsMover at interface for round 8
///               After every other rigid-body or glycosidic torsion angle perturbation perturbation
///                minimize entire complex
///                  MinMover (glycoligand backbone + sidechains + protein sidechains + docking interface
///             
///   Weights and Movers for rigid-body sampling
///    0.67 RigidBodyPerturbMover (rot = X, ang = Y)
///    0.33 the above and followed by FaDockingSlideIntoContact
///
///   Weights and Movers for glycosidic torsion angle sampling
///    0.45 Phi/Psi Sugar BB Sampling
///    0.3 Small BB Sampling - equal weight to phi, psi, or omega
///      -> 0.171429 +/- 15 degrees
///      -> 0.0857143 +/- 45 degrees
///      -> 0.0428571 +/- 90 degrees
///    0.2 Shear BB Sampling
///    0.0.5 RingPlaneFlipMover
///

Options and Scripting

Tips

See Working With Glycans for more information on carbohydrate modeling in Rosetta.

Typical Use

Here, the input prot_glyc_complex.pdb is a putative complex of a protein (chain A) and a glycoligand (chain X) where the lines defining the coordinates for each atom of the glycoligand is found at the bottom of the PDB file (this is to ensure proper setup of the FoldTree). The -refine_only flag is used to allow for only "small" perturbations of the glycosidic torsion angles during local docking. for The -auto_detect_glycan_connections flag tells Rosetta to determine the connections (i.e., the atomic bonds) between each carbohydrate residue of the glycoligand. In this way, Rosetta will create the necessary HETNAM and LINK records for the .pdb file.

/Rosetta/main/source/bin/./GlycanDock.default.macosclangrelease -include_sugars -auto_detect_glycan_connections -in:file:s prot_glyc_complex.pdb -nstruct 10 -carbohydrates:glycan_dock:refine_only true

See Also