• Make sure that you have a pdb-file for your ligand, with correctly added hydrogen atoms. This can be obtained with chimera or by Avogadro.
• It is also wise to rename atoms so that they follow a logical system at this stage, e.g. same numbers of H atoms and their parent C atom.

1. antechamber -i file -fi pdb -o hux.in -fo prepi -nc 1 -c bcc -rn HUX -at gaff
2. \rm ANTECH* ATOMTYPE.INF PREP.INF
3. Note that you often need to symmetrise some charges, e.g. of H atoms bound to the same C.
   

1. make_espinp.bash hux 1

   This will create two files: hux_opt that is a Gaussian-input file for geometry optimization using AM1, and q_esp_hux that is a Platon queue-script for executing geometry optimization followed by ESP calculation.

   • hux is the name of your ligand, and the script assumes that your pdb-file is named hux.pdb.
     
   • 1, the second argument to the script is the net charge of the ligand.
   • Be careful if the molecule contains Br or other atoms with two-letter code. The script sometimes give a wrong name in the Gaussian files.
     

2. Move the hux_opt and q_esp_hux files to Aurora. Submit the queue-script.
   On Kebnekaise, you need to change or add:
   #!/bin/bash
   #SBATCH -A SNIC2016-34-18
   module load gaussian/09.e.01-AVX
   
   If you want to keep dihedral angles fixed during the optimisation:
   Change the first line in the *_opt file to
   #P AM1 Opt=ModRedun
   and add a last line:
   * * * * F
   
   
3. The important output will be in the file hux_ff94.out. Move this file back to the local file system.
4. Type the following command:

   make_ambin.bash hux 1

   This will execute Antechamber and create one file: hux.in that is the Amber *.in and contains charges, atom types and topology.

   • hux is the name of your ligand.
   • 1, the second argument to the script is the net charge of the ligand.
   
   The command for Antechamber is:

   antechamber -i hux_ff94.out -fi gout -o hux.in -fo prepi -nc 1 -c resp -rn HUX -a hux.pdb -fa pdb -ao name

5. You can remove unnecessary output files with:
rm -f ATOMTYPE.INF PREP.INF QOUT esout qout punch ANTE* slurm-*.out sqm.in sqm.out sqm.pdb


6. Test the force field by running tleap and typing the following commands:

   source leaprc.gaff
loadamberprep hux.in
saveamberparm HUX prmtop prmcrd
quit


   If the force field works, you should not get any warnings starting with Could not find or No torsion.
   If you get these warnings, force field parameters are missing in GAFF and you need to create them, either by copying similar parameters from GAFF or performing frequency calculations.
   Note: the newest version of Amber contains much more parameters than the old one, so make sure that the newest leaprc.gaff is loaded.

   GAFF parameters are located in: /away/bio/AMBER/Amber11/dat/leap/parm/gaff.dat

7. Remove unnecessary files with: rm -f a.prm a.rst


8. Note that you often need to symmetrise some charges, e.g. of H atoms bound to the same C.
   

More manual procedure by Ulf

1. Optimise the molecule with QM methods, typically B3LYP/6-31G* or AM1 for a big molecule.
   
   
2. Calculate the electrostatic potential. For the Amber-99SB force field, this should be done at the HF/6-31G* level (B3LYP/6-31G* if you have metal sites). Other force fields require other levels of theory.
   
   Example of a Gaussian input file:
   %Chk=btn.chk
   #P B3LYP/6-31g* Opt
   
   Biotin, HF, 6-31G*, 1/9-05
   
      -1    1
   c      33.34700000000000    7.55100000000000   14.58000000000000
   ...
   h      36.15300000000000   11.00700000000000   16.90000000000000
   
   --Link1--
   %Chk=btn1.chk
   #P HF/6-31G* SCF=Tight Geom=AllCheck Guess=Read
      Pop=MK IOp(6/33=2,6/41=10,6/42=17)
   
   
3. Run antechamber on the output file of this calculation:
   
   antechamber -fi gout -fo prepi -c resp -i file.out -o res.in -rn RES -at amber -pf y
   
   where, you should insert the name of the output file from the previous point, and the residue name, (three letters) twice, first in lower-case and second in capital letters.
   Use -at amber, if want to use only standard Amber atom types. The default is to  get GAFF atom types, which is normally better
   With -pf y you will delete intermediate files.
   Alternatively, you can delete them by:
   \rm ANTECH* ATOMTYPE.INF PREP.INF
   
   
4. Often Amber gives you strange atom names. Then, you can use the program fixatomin to replace the wrong atom names. Insert the correct atom names into the file NEWPDB.PDB file and then read it into the program together with the NEWPDB.PDB file and the *.in file from antechamber.
   
   
5. Note that you often need to symmetrise some charges, e.g. of H atoms bound to the same C.
   
   
6. If antechamber fails, you may need to do the calculations by hand. Follow then the instructions below and on the RESP page.
   If you intend to use the Amber-02 polarisable force field, follow the instructions on the RESP page.
   

It defines the atom types and charges.
 How to make a new protonate entry
 How to define new parameters
 

Prep.in file

1. Start from a protonated pdb file, preferably with charges.
2. Run /teo/bio/Bin/changepdb, command Prep. This gives a file <residue_name>.in.
3. Change the title (row 3), if necessary.
4. Change the atom types (third column), if necessary. Appropriate atom types are found from similar amino acids or nucleic acids.
5. Change the tree symbol (fourth column). It defines the connectivity in the residue. Available options are:

M: main chain atoms (select this as the longest consecutive chain of atoms in the molecule).
E: end atoms (e.g. hydrogens). Not bound to any further atom.
S: side atom, connected to two other atoms.
B: branch atom, connected to three other atoms.
3: atom, connected to four other atoms (there are three branches out of this atom)
4-6: as for 3.
6. Change the charges (last column), if necessary. Take them from similar amino acids, from a resp calculation, or set them to zero (for ComQum).
7. Add loop closing statements, if there are any rings in the molecule. This is done with a the command LOOP two blank rows after the last atom and then a row for each ring with the name of the two atoms that close the ring.
8. Add improper torsions, if necessary.
   Normally we do not use them, but for ComQum link residues, they MUST be present in order for terms to cancel (i.e. if there are improper torsions in the mother residue, they must be present also in the truncated residue).
9. The resulting file is then read into leap with the command
   loadAmberPrep file.in
   

To avoid errors like in tleap:
+Currently only Sp3-Sp3/Sp3-Sp2/Sp2-Sp2 are supported
+---Tried to superimpose torsions for: *-C1-C6-*
+--- With Sp0 - Sp0
+--- Sp0 probably means a new atom type is involved
+--- which needs to be added via addAtomTypes

Issue the following command for all new parameters in tleap:
addAtomTypes {
  {"c1" "C" "sp2"}
  {"c2" "C" "sp2"} }

LOOP
C6  C1

DONE
STOP

1. Run $OML_AMBER/exe/prep -O -i <residue_name>.in

This gives <residue_name>.dat
Check that the total charge is correct.
You may also check if all bonds are coorect (this is done indirectly by edit later otherwise).
2. Copy the .in file to $OML_AMBER/dat/Mumod/Cprep

and the .dat file to $OML_AMBER/dat/Mumod.
cp *.in $OML_AMBER/dat/Mumod/Cprep
cp *.dat $OML_AMBER/dat/Mumod