If you're using Gromacs, you need the united-atom topology and coordinates. I would suggest using the "original geometry" file for coordinates, because otherwise ATB will have optimized the structure, which changes the coordinates and almost certainly will lead to clashes with the protein when you re-introduce the coordinates. To convert from .pdb to .gro, simply use editconf.

If you're using Gromacs, you need the united-atom topology and coordinates. I would suggest using the "original geometry" file for coordinates, because otherwise ATB will have optimized the structure, which changes the coordinates and almost certainly will lead to clashes with the protein when you re-introduce the coordinates. To convert from .pdb to .gro, simply use editconf.

Dear Justin,

Which software than can be used with ATP all atom topology for MD?

I don't know the purpose of all-atom files from ATB. The Gromos force fields are united-atom. Perhaps it is explained somewhere in the FAQ.

Here are some tips:

1) you don't need .gro file to start your simulation. you can easily use the original or optimized pdb file of your syste,. after finishing your first simulation (either em or md), gromacs will provide you with the .gro file.

2) For itp file, assuming you are using gromacs package, you should download files under GROMACS not the GROMOS. Don't wory about the all atom or united atom .itp as all required parameters for the bonded potentials is same in the GROMOS force field. You may only run into some problems with missing dihedral angles if you choose all atom force field.

I would tend to disagree with Abolfazi.

To point (1), while true that .gro files are not required (Gromacs happily accepts many formats), it can be very important which geometry you choose. If a docked ligand is then optimized in the absence of the receptor protein, the geometry can change substantially, leading to atomic clashes when one re-introduces the coordinates. This is very important in the case of crystallographic ligands, which generally do not have hydrogens added, so external software like ATB is necessary for building these on.

To point (2), it is very important to use only the united-atom topology! In looking at the all-atom files, it appears they are available to demonstrate how charges were summed to yield the united-atom topologies. The fact that methyl groups, for instance, use HC types for aliphatic H atoms is very concerning; HC is for aromatic H atoms only. Misapplying them can be very detrimental to the simulation. If there are missing parameters, that should set off alarm bells that something is wrong.

Thank you. I chose United-Atom PDB (original geometry) .

because I used GROMOS96 53a6 force field for my protein, I should choose coordinate and topology file from GROMOS96 Files. Is this the correct approach?

which coordinate file i should choose?, GROMOS96 United-Atom Topology or GROMOS96 United-Atom MTB ?

Dear Travis Pollard,

Is it right way if I choose GROMACS G53A6FF United-Atom (ITP file) for 53a6 force field in place of Gromos96 UA topology?

Thanks a lot !

thanks @Travis Pollard and @ Justin Lemkul

please help me about my problem in attached file

regards

Dear Justin Lemkul ,

I had passed a ligand containing 26 atoms through ATB server.

I have noticed that the ATB server retains only few hydrogens for United-atom topology creation. It is evident from the united atom (original geometry) pdb. (please see the attached lig_unitedatom.png). This is reflected even in the itp file as in this case only the details of 20 atoms are present.

However, in the case of all-atom files, all the atoms are considered. It retained all the atoms both in the all-atom original geometry pdb and .itp files. (attached as lig_allatom.png).

In my humble opinion, using United-atom files might give artificial interactions which are not relevant as few of the atoms are missing.

Request you to give your opinion on this.

Thank you,

Regards.

GROMOS is a united-atom force field. This is its convention. If you believe an all-atom force field is more suitable, you'll need to use the appropriate tools for it.

Dear Justin Lemkul,

Thank you for the inputs.