A.- In-house method (using Gaussian)
1.- Get a PDB of the molecule.- NOTE: be careful when naming the atoms; try to use the same nomenclature that has been already accepted/used for this kind of substrate. In addition, order it in a logical way based on the atom connectivities.
2.-Prepare the Z-matrix.- Amongst the different input formats that are supported by Gaussian, we use to define the coordinates as a Z-matrix using Molden or Babel. The former allow you to check or redifine the Z-matrix.
3.-Optimize the geometry using HF/3-21G as follows:
#RHF/3-21G opt=Z-matrix scf=direct
Molecule geometry optimization
0 1 (charge multiplicity)
4.-We calculate the charge distribution using HF/6-31G* and reading the *chk file from step 3 as follows:
#RHF/6-31G* geom=check guess=read pop=mk IOP(6/33=2)
0 1 (charge multiplicity)
5.-Finally, using the last *log from Gaussian we generate the PREP file using ANTECHAMBER (AmberTools):
antechamber -fi gout -i myfile.log -fo prepi -o myfile.prp -at amber/gaff -pf yes -c resp
at ::= atom type. If your substrate is an amino acid-like maybe is better to use AMBER type; if not, use GAFF. But do not forget to include gaff.dat parameters (loadamberparams gaff.dat) in your tleap session.
pf ::= if you indicate yes (recommended), only the final files generated by Antechember will be stored.
(more options antechamber –h)
'NOTE: once the *prp file has been generated, change the 3-digits code at the head of the file (e.g., if you are calculating the parameters for leucine it is bound you will use LEU).'
6.- Generate the force field parameter file for your molecule using the program parmchk (AmberTools).
parmchk -i myfile.prp -o myfile.frcmod -f prepi
And that's all!
B.- Online method using R.E.D. Server (also using Gaussian)
Working flow: PDB –> p2n file (input for REDS) –> output
1.- Edit your PDB file. Atoms must be ordered by their connectivity.
2.- Go to R.E.D. server > Submit > Private account (recommended; your projects are stored up to 2 months) > Use Ante_R.E.D. 2.0 > Non-automatic mode…
There you will find the following options:
-Scaling Factor (1.2 default). The worst the input geometry, the larger the scaling factor value that has to be used. In our case (PDBs generated using PyMOL, Corina3D or Molden 1.2 is OK).
-Correct the atom names (only redundant atom name by default). In case you have redundant atom names, AnteRED will change it.
-Reorder atoms (approach ONE by default). I always deactivate this option because I have been really careful checking the connectivities in PyMOL and Molden.
-Root for connectivities. You can define with heavy atom will be the root for the connectivity of your system. Take one atom of the terminal site of the molecule.
-Number of residues. For single molecules, 0.0 is recommended.
-Number of PDB file(s). You can upload n-PDBs to generate n-p2n files.
Pressing Next, you could load the PDB files. Finally, Let's go.
3.- In a relative short period of time (ca. 5 s) you will have available the p2n file. I copy here an example for 4.pdb:
REMARK 1 P2N generated by Ante_R.E.D. version 2.0
REMARK TITLE MOLECULE
REMARK CHARGE-VALUE 0
REMARK MULTIPLICITY-VALUE 1
REMARK REORIENT 1 6 3 | 3 6 1
ATOM 1 O1 ETF 1 42.139 44.841 43.848 O O
ATOM 2 H1 ETF 1 41.827 44.021 43.404 HO H
ATOM 3 CT3 ETF 1 39.759 45.175 44.186 CA C
ATOM 4 H2 ETF 1 41.331 46.668 44.370 HA2 H
ATOM 5 H2 ETF 1 41.021 46.146 42.711 HA3 H
ATOM 6 CT2 ETF 1 41.091 45.806 43.745 CB C
ATOM 7 H3 ETF 1 39.829 44.849 45.224 HB2 H
ATOM 8 H3 ETF 1 39.522 44.320 43.553 HB3 H
ATOM 9 CT4 ETF 1 38.632 46.218 44.070 CG2 C
ATOM 10 H4 ETF 1 38.357 46.341 43.022 HG2 H
ATOM 11 H4 ETF 1 38.975 47.171 44.471 HG3 H
ATOM 12 CT5 ETF 1 37.406 45.738 44.868 CD3 C
ATOM 13 H5 ETF 1 37.169 46.465 45.644 HD2 H
ATOM 14 H5 ETF 1 37.625 44.774 45.328 HD3 H
ATOM 15 H5 ETF 1 36.554 45.633 44.196 HD4 H
CONECT 1 2 6
CONECT 2 1
CONECT 3 6 7 8 9
CONECT 4 6
CONECT 5 6
CONECT 6 1 3 4 5
CONECT 7 3
CONECT 8 3
CONECT 9 3 10 11 12
CONECT 10 9
CONECT 11 9
CONECT 12 9 13 14 15
CONECT 13 12
CONECT 14 12
CONECT 15 12
In the head part of the file you define:
- Title of the molecule - Charge - Multiplicity - Reorient. Related to the reorientation algorithm (improvement vs. method A commented above). More info: http://q4md-forcefieldtools.org/Tutorial/Tutorial-1.php#REORIENT - And coordinates & connectivities.
(NOTE: there are other REMARKS really useful for amino acids et al. I do not include them in here).
4.- Submit your P2N file using RED IV (Go to R.E.D. server > Submit > Private account (recommended; your projects are stored up to 2 months) > Use RED IV
Choose Mode 1 unless your system has already been optimized by QM (this is not our case).
- Number of molecules - Charge model (I use RESP-A1A; HF/6-31G*HF/6-31G* - Connolly surface algo. used in MEP computation - 2 stage RESP fit qwt=0.0005/0.001 - Quantum mechanics program: Gaussian_09 Finally, next and wait for the end of the calculation (ca. 2 h for a 20 atom system) 5.- Once the calculation is done, download the folder with your results from the web site. Then, decompress it, go into Data-** folder and use Antechamber to generate from the *mol2 file (it contain the optimized geometry and charges): antechamber -fi mol2 -i Mol_m1-o1-sm.mol2 -fo prepi -o myfile.prp -at amber/gaff -pf yes (NOTE: there are 2 or 3 *mol2 file. Select the last one, Mol_m1-o1-sm.mol2). Again, if you need to generate force field parameters for your molecules, used parmchk. Hope it helps!