Generating topology for charged molecules

Question

I have to generate a topology file for a molecule like Sodium Triphosphate, and I do not find forcefields available in the literature. So as a starting point would like to use engines like LigParGen for OPLS. I will have to submit one molecule without the sodium ions.

• As it accepts molecules only upto a charge of -2 and I have a charge of -5, can I consider adding hydrogens to terminal charged oxygens, generate the file, delete the hydrogens and all data related to them under all directives, and add the extra charge back onto the oxygens?
• Why or why not is this procedure correct?
• Is there a better way to approach this problem in case of polymers with a lot of charge?

Answer 1

There are two options here and the choice depends on what you intend to do with your results.

1. If you are going to perform QM calculations after MD, then you can just alter the charges as you outlined in your question. This is because QM will correct any issues you might come across. However, you need to make sure the molecule will behave in the manner in which you would expect. You cannot just cram some parameters in a molecule to make the code accept them.

2. If you are only doing MD, you might be able to borrow the parameters or follow the procedure outlined in this paper. They use a system very similar to yours.

Even if you are performing QM calculations afterwards, you could just use procedure 2 in general.