I have used ATB (http://compbio.chemistry.uq.edu.au/atb/) server to generate topology for piroxicam molecule (having 36 atoms). It has generated topology for GROMOS96 force field with two charge groups (one contains 2 atoms and the other 34 atoms). Is it reasonable? I am using GROMACS for MD simulation. It cannot process a charge group containing more than 32 atoms. What can be the way-out?
You could modify them by hand. Create "reasonable" charge groups with just a few atoms like one for each functional group, one for the amide bond etc.
Also, you may find valuable hints in
http://pubs.acs.org/doi/abs/10.1021/ci100335w
aminoacids.rtp file contains the topology of all the amino acids. More than one amino acid may have the same functional group. How will I choose the conistent functional group from the residue topology file to create a reasonable charge group?
Sorry, I'm not sure what you mean.
The charge groups do not have to be unique over all defined building blocks (i.e. residue definitions). Every residue topology has its own assigned charge groups, independently from other residues. As long as you keep the number of atoms per charge group low and the overall charge sums to an integer value, you should be fine. Furthermore, if I remember correctly, the influence of charge groups is very small when using PME. But better check the manual (page 24 of 4.6.3) on this point.
Good luck.
Say, amide bond of piroxicam is one charge group. I have to assign partial charges to N-atom, H-atom etc such that total charge is 0. Now, what would be the partial charges of these atoms? As piroxicam is not a part of the GROMOS96 force field, partial charge distribution is not found in the topology file. I will see the amide bond in the existing topology (which is residue topology of amino acids) in the force field directory and look for the partial charge distributions among N-atom, H-atom etc. But amide bond is present in many residues . Now, which residue I will follow to assign partial charges among the atoms in the amide bond of piroxicam?
I would use the charges which the ATB server calculated. You can adjust them slightly, if it is necessary to get an integer value.
Looking at your .itp file, I would suggest two more things: Firstly, why do you use the all-atom version? You are going to use the united atom force field Gromos, which omits all the aliphatic hydrogens (i.e. atom type "HC"). Use the united-atom version of the itp provided by the server and you will have less atoms, which will make the charge group thing easier. Secondly, the solvent atom type "SDmso" was assigned to atom number 17 (sulfur). Maybe you should consider changing it to "S".
Unfortunately, united atom force field topology also contains two charge groups (one contains 2 atoms and the other 31 atoms). I will see whether I can have more charge groups by changing tha value slightly. Thanks.
- Badges
- Science topic