Zn-Cys site, covalent ligand, ATP/Mg2+, and Gaussian setup for Br compounds

[rajiv03]

Thanks for the great tool. I would like to ask if these functions are available now. If not, it would be nice to have them in the future. Thanks!

Can Steamd handle a simulation of a zinc-bound cysteine in a protein that also has a ligand bound at another site? I tried specifying the protein and providing Zn as a cofactor, but it doesn't work properly. Are there any specific steps or parameters I need to set up this type of simulation correctly?
Can Steamd handle simulations involving a covalently bound ligand inhibitor in a protein? Are there any special considerations or setup required to model covalent bonds accurately?
How can I set up a Steamd simulation of a protein-ligand complex in the presence of ATP and Mg2+ (or other ions)? What is the recommended way to configure this type of simulation to properly account for the interactions between the protein, ligand, ATP, and ions?
Are there any specific installation instructions or details I should be aware of for using Gaussian software on Linux to model compounds containing bromine?

Thanks!

[avnikonenko]

Hello!
Thanks for your interest in the tool!

1. Simulations with zinc-bound cysteine and a ligand.
   To run simulations involving ZN and a ligand with MCPB.py, provide a regular protein PDB file containing the ZN atom -p protein_ZN.pdb and separate ligand.mol/sdf files '-l ligand.mol'. Also to let StreaMD know that you want to treat active site related to ZN atom by MCPB.py procedure you should provide specific arguments to activate Gaussian software --activate_gaussian "module load Gaussian" --gaussian_exe g09 (specific Gaussian version and activation string should be your own machine related). Provide the metal residue name to be processed by MCPB.py --metal_resnames ZN.
   Additionally, you can specify a few other arguments such as --metal_cutoff by default, 2.8 and --metal_charges by default, {MN:2, ZN:2, CA:2}.
   More information about the procedure you can find in the MCPB.py article and tutorial.
   StreaMD is only a wrapper over the actual MCPB.py tool, to help users with technical aspects. Unfortunately, there are too many steps which can go wrong and also it is very highly system dependent. During my own experience there can be incorrect choice of basis functions, wrong protonation states, or starting geometries with convergence problems. Therefore, despite the technical aspects of the pipeline functioning correctly, it still would require knowledge and experience from you to manually identify the system-dependent source of the problem. So I would recommend firstly to make sure you have an information about the method itself. If any error occurs you can take a look into StreaMD logging files and try to find the source of it.
2. StreaMD can not correctly simulate systems involving a covalently bound ligand. It requires additional parameters for covalently modified amino acids, which are not presented in the default Gromacs supported force fields. Unfortunately, I don't have enough expert knowledge about such simulations.
3. Ligand, ATP, Mg2+ simulation.
   AMBER force fields (By default, StreaMD uses AMBER99SB-ILDN protein, nucleic AMBER94 (Lindorff-Larsen et al., Proteins 78, 1950-58, 2010)) support some ions, including Mg2+. So you can run such simulations by regular (without Gaussian) Gromacs run. Specify ligand and ATP as separate mol files, while keeping Mg2+ in the protein.pdb file.
   run_md -p protein_Mg.pdb -l ligand.mol --cofactor ATP.mol .
   Also you still can use Gaussian and MCPB.py for more precise parameterisation of Mg2+ active site if output simulation will look incorrect (distorsion in the geometry) by using default Amber force field.
4. Regarding Gaussian software, the Gaussian license is required and must be held by the user. Although, Gaussian parameterisation is needed only for non-standard systems (Metalloproteins containing metal atoms which are not supported by standard FFs or non-standard ligands, such as boron-containing).
   As for compounds containing bromine they can be prepared by standard Antechamber procedure and don't require Gaussian.
   Hope it is helpful and let me know if you have any other questions!