Hello everyone.
What would be an appropriate procedure to validate parameterization of a ligand using a general small molecule forcefield (such as General Amber Forcefield, CHARMM General Forcefield, OpenForceField 1.0.0 Parsley)?
Is it enough to check the minimized structure against some high level QM calculations?
Thanks a lot
Parameter validation can become quite challenging, based on the work you intend to do and the results you expect to achieve.
You can take a look to this tutorial for some initial checks:
http://www.ub.edu/bl/2017/03/16/small-molecule-dihedrals-parametrization/
Also you may take a look to the following papers:
10.1021/acs.jcim.9b00552
10.1021/acs.jctc.8b01039
10.1021/acs.jcim.7b00221
10.1021/acs.jcim.7b00505
Hope this helps
Hi Heiner,
One option could be that you get reproduce an experimental protein-ligand binding mode, after submitting your protein-ligand system to molecular dynamics studies.
Hi, you shall validate towards experimental data by running MD simulations for your compounds. For example, if you found density at a certain temperature for your new compound you can make a system and run MD to see if you can reproduce the experimental density. Then you might even find enthalpies of vaporisation and run free energy calculations in order to see if you can reproduce them.
Thank you for your suggestions Fernando Prieto-Martínez , Martiniano Bello , Inna Ermilova
As I am working on covalent ligands, reproducing an experimental binding pose is unfortunately not an option (only structures for the covalent adduct have been resolved so far).
What about comparing MM frequencies to QM vibrational calculations? I guess after already having used resources for the QM calculations I could directly derive parameters from there ... but nonetheless, would you consider this an appropriate approach?
If you have access to QM calculations it is an excellent starting point, just remember to use similar basis sets to the ones used for the parametrization of the forcefield you intend to use. There are numerous sources which discuss similar cases, e.g., parametrization of zinc proteins or MOFs. With that said an alternative approach could be QM/MM in you complex.
Best regards
Dear Heiner, I know that in many cases QM calculations can be of consideration as well, particularly, when there is no experiment was done or could be done. For instance, if similar methodolgy was used for deriving other parameters accuratly, then you could motivate using experience of previous developers.
- Badges
- Science topic