Your question is not very clear.

A "bimolecular reaction" is probably best carried out in QM software. MM programs like CHARMM, while they have interfaces to QM software, do not directly do QM calculations like chemical reactions.

Are you confused about the format of the RTF or the actual parametrization process? The former is rather simple, as a residue is simply defined by its atoms, their charges and types, and their connectivity. Parametrizing a new molecule is not simple but can be greatly aided by the CGenFF program (there is a web server that will do everything you need, unless the topology needs considerable revision, in which case you must follow the tutorial).

Thank you very much for your quick response Justin Lemkul.

 I am confused about the fact that CHARMM has some pre-saved RTF files for some proteins. But if I have to create a new RTF file there are terms like TIPS3 and the connectivity, how do we decide upon that? 

I wanted to know if I can get an example of RTF file of  bimolecules other than proteins ?

There are many RTF files distributed as part of the full CHARMM force field (nucleic acids, lipids, carbohydrates, other small molecules, etc), and stream files with many other molecules. Without a more specific question (why you need to create a new RTF file, what its contents are, etc) it's hard to be more helpful.

Hello Justin Lemkul

Thank you once again for your quick response. Actually I am trying to study a bimolecular interaction energy (VdW ) for molecules other than proteins , carbohydrates , lipids etc. So, I was guessing if I have to generate a RTF and a  Parameter file using CGenFF?

Yes, if you have some novel molecules that are not covered by the force field, you need to derive its parameters. The process for doing so depends a little bit on what those molecules are, but CGenFF is a good place to start.  Note that refinement is often necessary, so you need to be familiar with the parametrization procedure in order to produce high-quality topologies.

Justin Lemkul

Ok, Thank you very much for your precious guidance. 

 Justin

I am really confused. I am a beginner and even after reading through the available material online I couldn't really understand the exact procedure I need to follow to calculate the vanderwaal interaction energy between two molecules. I am clear with the fact that I have to generate RTF and Param file first. I am using an online software "https://cgenff.paramchem.org/initguess/"  that will help me to generate my stream file. But, what I have to do after that? I would be really grateful, if you could guide me with the right path.

Thanks

The CGenFF server will give you a topology.  It won't give you energies. The evaluation of that topology needs to be against some well-defined target data, but the stream file you get gives you an idea of how good the topology is. High penalties mean you need to refine the parameters; this may be above your skill level so you should get some help from your advisor or whoever else is available.  The CGenFF paper is extremely detailed in how to do all of this.

If you want to study interactions between molecules, that's what actual simulations (using the generated topology) are for.

Hello Justin

Thank you very much for your response. I am now getting the path to be followed.

I have a well defined structure that I need to study and develop my topology file for and using Babel I converted it into pdb format and ran it on CGenFF. I got the stream file which had both RTF card and Param card. But the penalties were quite high (144). This I guess is because of the change from (QM calculation from which my structure is developed) to  force field (MM calculation).

I was wondering, since I am evaluating the van der waal energy, do I really need to be concerned about the penalties because my focus is on evaluating the non bonded energies?

As the saying goes, "garbage in, garbage out." The results of your simulations are only as good as the input parameters. If you have large penalties, this means the analogy between existing CGenFF parameters and those required by your molecule is poor in those cases. Those are instances where you need to refine the parameters. If the charges have large penalties, you certainly cannot leave them alone as the balance between nonbonded interactions will be poor and any results you get will have no value. Charges are refined by targeting QM dipole moments and then slightly overestimating these when redistributing charge.  If the bonded parameters have large penalties (especially important for dihedrals, which control conformational sampling, and therefore dictate how intramolecular interactions behave, which in turn have implications for intermolecular interactions) then you need to do QM potential energy scans and refine the parameters until you get agreement with the relative conformational energies.  All of this is in the CGenFF tutorial available online. 

Parametrization is hard work. Even adding some "simple" molecule to a force field may take considerable effort.

Hello Justin Lemkul

Apologies for the late response! 

Thank you very very much for this guidance. It is very clear ,detailed and helpful to step forward. I highly appreciate your patience and guidance.

Many Thanks and Regards

Hello Justin Lemkul

Apologies for the late response! 

Thank you very very much for this guidance. It is very clear ,detailed and helpful to step forward. I highly appreciate your patience and guidance.

Many Thanks and Regards