Now that the electrostatic potential is known we need to use this to produce a ligand description for incorporation into the Amber force field.

RESP fitting

Once the electrostatic potential has been successfully generated a two step process is used to convert this into RESP fitted charges. In these commands the new options are; -c charge method and cf for charge filename. The '.ac' file is an intermediate file and most options are select explanatory, except -c wc which simply means 'write out charge'.

antechamber -fi gout -fo ac -i 4bjx-ligand.out -o resp.ac -c resp
antechamber -fi ac -i resp.ac -c wc -cf resp.crg

These commands will produce the files resp.ac and resp.crg, as well as may temporary intermediate files (most with names beginning with ANTECHAMBER).

Create prep and frcmod files

The goal of the parameterization is to create two files used by Amber to read a parameterized drug. The first is a .prep file which adds a new residue to the standard Amber residue database. The other is a force field parameter modification, or .frcmod, file which lists parameters for bonds and other components of the forcefield that need to be added to characterize the ligand.

At this stage you need to consider a name for the ligand to be used in the simulation. The residue name must be 3 characters long, in this example we use 'L01'. BAC requires that this name in lowercase letters must be used consistently throughout the rest of the process to name all files associated with the ligand (for instance the l01.prep file generated below).

antechamber -i resp.ac -fi ac -c rc -cf resp.crg -o l01.prep -fo prepi -ao name -a 4bjx-ligand-h.pdb -fa pdb -rn L01

In this command, -a sets an additional file to be read for naming, -ao name means that only atom names are read from the input (rather than also including atom and bond types) and -rn L01 sets the residue name.

Now we have the .prep file we need to use Antechamber to try and fill in any missing parameters in the force field by analogy to similar parameters which are present. If Antechamber cannot empirically calculate a value, or has no analogy, it will either add a default value that it thinks is reasonable or , in rare cases, insert a place holder. This will contain zeros in all columns and be accompanied by the comment "ATTN: needs revision" in the output. In these circumstances you will have to manually add these parameters.

parmchk -i l01.prep -f prepi -o l01.frcmod

Finally we need to produce a PDB which contains the ligand labelled using the new name we have given it. An easy way of achieving this is to run the command:

sed 's/73B/L01/' 4bjx-ligand-h.pdb > l01.pdb

Remove any header lines from the resultant PDB (reduce adds one beginning 'USER MOD').

For use in BAC the ligand PDB ( l01.pdb in the example) should be checked to ensure that it does not contain entries in the element column. In particular entries for chlorine can be a problem.

Non-standard approach (using AM1-BCC)

This section is something of an aside and is included for completeness only, skip if you are only interested in creating standard BAC parameterizations. If you wish to create the topology without Gaussian optimization and using the AM1-BCC approach to charges this can be achieved with a single command:

antechamber -i 4bjx-ligand-h.pdb -fi pdb -o l01.prep -fo prepi -c bcc -at gaff -j 4 -rn L01

Here the ligand is renamed 'L01' as in the Gaussian method.

Test the parameters

To test that this process has gone correctly it is often a good idea to build a ligand only system which could be used for simulation. To do so we need to use a tLeap script, like the one below:

source leaprc.gaff
frcmod = loadamberparams l01.frcmod
loadamberprep l01.prep
mol = loadpdb l01.pdb
saveamberprep mol test.prepc
saveamberparm mol test.top test.crd
savepdb mol test.pdb
quit

This script loads the GAFF force field and the files we have created to describe our ligand. It then used these to create simulation ready topology and coordinate files. A copy of this script can be downloaded here: test.leapin. Run this to produce a ligand model using tleap:

tleap -f test.leapin

A common reason for failure is that some atom names change capitalization in the Antechamber process (in particular 'CL' becoming 'Cl' and 'BR' turning into 'Br'). It is easiest to change the case in the PDB as the entries only appear once (as opposed to the .prep file where they appear in atom and connectivity lines). If one is not present you should add a line containing only the text TER to the end of the ligand PDB. A corrected PDB for our ligand can be downloaded here.