I am studying a Diels-Alder reaction of cyclopentadiene and maleic anhydride. I know that AMBER can be used to run MD simulations on proteins, however would I be able to use these small molecules to run a simulation?
$\begingroup$
$\endgroup$
2
-
$\begingroup$ Contrary to what folks are saying, it all depends on if you want thermodynamics or kinetics. If you are after thermodynamic properties, all you need is the starting and ending states, and you sure the hell can simulate these with classical forcefields, and alot of folks do, for reactions! If you are after kinetics, you would not be able to use classical force-fields. $\endgroup$– B. KellyApr 20 at 3:14
-
$\begingroup$ @B.Kelly Fair point, though you will need more data if you want to calculate free energy difference or something similar. $\endgroup$– GregApr 20 at 13:18
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
Hemanth has implied it, but I think it is better to say more directly: AMBER and other classical forcefields cannot simulate bond breaking or reaction. If you stretch a bond in these simulations, however far you stretch them, the atoms stay bonded in the original way and return to the original molecular geometry when you let them go.
If you want an MD simulation that allows reactions and bond breaking, you need a reactive forcefield (ReaxFF) or simulation based on a theory that includes electrons, etc (DFT, DFTB etc)
$\begingroup$
$\endgroup$
Short Answer No. To elaborate, AMBER or CHARMM or any of the Classical force fields are employed for equilibrium MD simulations. The FF description would generally follow the format of energy terms of bonds, then angles and dihedrals and other parameters. These are parameterized for their equilibrium values and are represented as harmonic springs. In your particular use case, you should ideally consider a force field like ReaxFF which is employed to study reactions.
$\begingroup$
$\endgroup$
As previously mentioned, you can not use classical molecular dynamics codes to study reactions.
Faster than running MD based on DFT, you can try the xTB software, where you will have at least three different semiempirical Hamiltonian to be used with the MD calculations.
