Query: Running simulation with external parameters

[CallumFrEllis]

Hi,

I am trying to run some small simulations usings the parameters listed in the SIs for https://pubs.acs.org/doi/10.1021/jp410720y for a variety of organic solvents.

What is is the best way to use these parameters? Can I create a key file with these, or append them to the Amoeba09.prm or create a new .prm file with only these parameters?

Any help would be appriciated.

Best wishes,

Callum

[jayponder]

Actually, you can do any of the three things you suggest.

Any "extra" parameters needed for the system you are computing can always be added to the key file. In that case you would still use a line in the key file pointing to an AMOEBA parameter file (for example "parameters /Users/ponder/tinker/params/amoeba09.prm" or similar). Parameters found in the key file are read after any main parameter file such as amoeba09.prm. And thus parameters in the key file with the same atom type or class numbers as in the parameter file will be overwritten by the values supplied in the key file.

Alternatively, you could simply add the same extra or new parameters directly into the amoeba09.prm or similar parameter file.

Finally, you could create your own new parameter file. If you choose to do this, then make sure you copy the "header" information for the AMOEBA force field into your new parameter file. This information can be found as the top section of any of the AMOEBA parameter files, and serves to define the energy functions used by the AMOEBA force field.

Hope this helps. Let me know if anything is unclear.

[CallumFrEllis]

Thank you so much for the prompt and very helpful response, I will have a play and see how I get on.

Thanks again,

Callum

[CallumFrEllis]

Hi Prof Ponder,

I have another potential straightforward question, I am trying to create n-alpha acetylated methyl ester glutamine, which I believe I can do with the atom types available between amoebapro13.prm and amoeba09.prm with the ace capped GLN parameters from amoebapro13.prm and then the methyl ether parameters from amoeba09.prm. Looking to append the methyl ether atoms to the end of the amoebapro13.prm file. Would this be correct or will I have to find a way to parameterise this molecule as a whole? Or is there another way of doing this that might be easier- just putting the two seperate parameter files in the .key file? I have renumbered the methyl ether atoms but get READXYZ- check connection of atom errors.

Thanks,

Callum

[jayponder]

Due to the fact that AMOEBA uses polarizable multipoles and strives for a highly accurate representation of electrostatics, it is not as simple to "mix-and-match" parameters for AMOEBA as it is for simple partial charge force fields. Parameters generated this way will not be as accurate as the usual standard for AMOEBA. That said, your case is relatively simple and you might be able to "merge" values from the AMOEBAPRO13 and AMOEBA09 parameter files. I'd suggest using doubly capped ACE-GLN-NME from the protein model, and then replace -NHMe with -OMe. Of course, as with a simple partial charge model, you will need to at least make sure the sum of the atomic charges in your final model is zero.

Alternatively, and more generally, you could generate parameters for the whole ACE-GLN-OMe molecule using the Poltype automated parameterization system. Poltype is available from the TinkerTools organization site on GitHub. If your project is limited to just this one specific molecule, you might first try the "merging" method. If you envision doing several other systems in the future, then I'd suggest you investigate using Poltype. Poltype is not completely trivial to get working and has something of a learning curve, so its use requires some time and effort.

[CallumFrEllis]

I am endeavouring to use the 'merge' method, but I know I need to be accurate with the atom types/classes. I have created a 'new' .prm file for my key file, but I can't see the atom classes for the amoeba09.prm set as is outlined at the top of the amoebapro13.prm file so I obviously get undefined parameter errors.
Forgive my ignorance but where can I find the defined atom classes for amoeba09.prm? The atoms in question are class 36, 40, and 44, which obviously do not line up with the ones defined in ameobapro13.prm. Thanks for any help you could provide.

Callum