ForceField - Set force field

CAP - Catabolite gene Activating Protein (1BER)

Endonuclease PvuII bound to palindromic DNA recognition site CAGCTG (1PVI)

GCN4 - leucine zipper transcription factor bound to palindromic DNA recognition site ATGAC(G)TCAT (1YSA)

°

-

Set force field

Command

Argument

Datatype

Default

Min

Max

Format:

ForceField

Name = Force field name,

STRING

Method = AutoSMILES | AM1BCC | AM1,

STRING

AutoSMILES

SetPar = Yes | No

STRING

Python:

ForceField(name,method=None,setpar=None)

Menu:

Simulation > Force field

Longrange, Interactions, Cutoff , ScaleForce, SimSpeed

Required:

The ForceField command sets the force field used for molecular dynamics simulations. Currently, the following force fields are available by default: Amber94, Amber96, Amber99, NOVA, Yamber, Yamber2 and Yamber3.

The force field parameters are stored in *.fof files in subdirectory fof. After changing or adding parameters there, the force fields must be updated by running YASARA with the -upd command line parameter. This will recreate the file yasara.fof, which contains all force fields in compressed form.

The actual force field parameter assignment takes place at a later stage, whenever the force field is initialized.

The Method parameter chooses the approach used for point charge assignment. 'AutoSMILES', an extension of AM1BCC, is the default. Selecting 'AM1BCC' or even 'AM1' is mainly required to compare results between different methods, 'AM1' charges are not suitable for molecular dynamics simulation.

It depends on the application which force field gives the best results: (More details in be found in the corresponding force field references listed in the citation section)

The NOVA force field has been optimized for energy minimizations of proteins in vacuo, the force cutoff distance should be set to 10.48 A. The main problem with force fields used in vacuo is that the strength of electrostatic interactions is heavily overestimated, since shielding water molecules are absent. One solution is to reduce the net charge of charged amino acids like Lys or Glu. Some force fields (e.g. GROMOS) set it to 0, while NOVA choses 0.23, a value that has been optimized together with the other NOVA parameters.

The AMBER force fields have been developed for general use in all kinds of applications, including simulations of nucleic acids. The force field is meant for simulations in explicit solvent. When used in vacuo for quick modeling tasks, electrostatic interactions (like formation of salt-bridges) will be more pronounced than in reality.

The Yamber force fields have been optimized for stable simulations of proteins, maintainance of structural quality and improvement of homology models. High resolution X-ray structures simulated with Yamber usually stay closer to the native conformation and show a smaller deterioration of knowledge-based quality indicators (Ramachandran plot etc.) than the other force fields. For other applications than the ones listed (e.g. studies of protein dynamics and nucleic acids) a simulation with the classic AMBER force field should be run for comparison. All Yamber force fields have been validated with a 7.86 A force cutoff, long range electrostatic interactions using the PME algorithm and explicit solvent.

Yamber uses the same atom types and equilibrium bond lengths/angles as AMBER, while Yamber2 employs the Engh&Huber parameters commonly used for X-ray structure refinement and structure validation. Yamber3 is similar to Yamber2, just that 20 CPUs spent an additional year on finding the optimal parameters. It should thus perform a bit better than Yamber2, but is conceptually so similar that the original Yamber/Yamber2 reference can be cited.

As described above, every force field has its own optimal cutoff and electrostatics treatment , which is summarized in the table below. The SetPar parameter allows to set these parameters automatically together with the force field, obviating the need to set them separately using the Cutoff , Longrange and Boundary commands.

Optimal simulation parameters:


Force field	Cutoff	Longrange	Boundary
NOVA	10.48	None	Wall
AMBER	?	Coulomb	Periodic
YAMBER	7.86	Coulomb	Periodic
YASARA	7.86	Coulomb	Periodic

The treatment of metal binding using only Coulomb and Van der Waals interactions of classical molecular dynamics force fields is difficult, i.e. sometimes experimental distances are not exactly reproduced. Distances to Asp and Glu tend to be shorter due to electrostatic attraction, while distances to His are usually longer. Normally results are accurate enough to allow for stable metal binding during the entire simulation, but for complex metal interactions like Porphyrin rings, it is helpful to fix the distance with pseudo-bonds.

The treatment of dummy atoms (chemical element 0, named 'Du') differs between force fields: in AMBER-like force fields they do not exert any force at all, while the NOVA force fields treats them like a single hydrogen with charge 0.

Today's semi-empirical force fields suffer from a number of limitations , among which the existence of energy traps at short distance is usually not well known, but can become important when energy minimizing structures with lots of bumps:

AMBER-like force fields (including YAMBER/YASARA) contain polar hydrogens (water and hydroxyl groups) without repulsive Lennard-Jones interactions. Since these hydrogens are positively charged, they will be attracted by negatively charged atoms until either their covalently bound oxygen stops the approach via its own Lennard-Jones repulsion, or until nuclear fusion occurs. The latter may happen if the polar hydrogen is already placed inside another atom at the beginning of an energy minimization. The solution is to disable the electrostatic forces during an initial steepest descent minimization, best restricted to the side-chain atoms. This is done automatically by the energy minimization experiment.

The NOVA force field enforces planarity by fitting a least-squares plane through each planar group and applying forces towards this plane. If initial bumps cause a planar group to bend by more than 90 degrees, it may get trapped in the bent conformation. The solution is to disable planarity forces during an initial steepest descent minimization, best restricted to the side-chain atoms. This is also done automatically by the energy minimization experiment.

Example 1:

ForceField Amber99

Choose Amber99 force field for simulations in solution.

Example 2:

ForceField Nova,SetPar=Yes

Choose Nova force field for energy minimizations in vacuo and set default parameters.

Example 3:

ForceField Yamber3,Method=AM1BCC

Choose Yamber3 combined with standard AM1BCC charged for newly parameterized molecules.

Example macro:


# EXAMPLE ForceField
# Requires YASARA Dynamics
Clear
# Load myoglobin
LoadPDB 1a6g
# Prepare for simulation
Clean
# Style it
Style BallStick
Style Ribbon
ShowAtom SideChain CA Res His 93
PosObj 1,X=4,Y=1,Z=10
OriObj 1,Alpha=-68,Beta=40,Gamma=75
# Create cell
Cell Auto
Boundary Periodic
# Set force field
ForceField Amber96
Cutoff 7.86
LongRange Coulomb
# Start simulation
TempCtrl SteepDes
Sim On

Figure: Result of the example macro above.