Preparing the ligand

From DISI
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Preparing a ligand

Automatic way, starting from SMILES

This way, you will make use of John's automatic scripts for database preparation and actually upload new molecules to a special section of ZINC.

  • it is advisable to create a special subdirectory, since many new files will be generated.
  • the file containing the SMILES strings should contain a string followed by an identifier on each line.
  • OPTIONAL: run convert.py --i=yourname.smi --o=yourname.ism . This will convert your SMILES to isomeric SMILES.
  • run dbgen.csh yourname.smi.
    • Note that the dbgen.csh does not work for more that 1000 molecules.
    • Brake up your molecules into chunks of 1000 and run dbgen on each chunk.
    • Clean up your directory afterwards. dbgen.csh generates a lot of files that you do not need if it ran correctly.
  • you should obtain a file somename.db.gz .

optional

To increase the number of molecules that are written out for the database generation, copy the file $DOCK_BASE/data/omega.parm into the directory that dbgen.csh is going to be run in.

  • At the end of the omega.parm file you will see a section called "Torsion Driving Parameters", here you will find three variables that can be changed.
      • SetMaxConfs(600) #set to higher numbers ie. 1000
      • SetRMSThreshold(0.80) #set to lower numbers ie. 0.50
      • SetEnergyWindow(12.5) #can be changed but this can often generate broken molecules
  • WARNING this should only be done if generating conformations for a small set of compounds!!!

Manual way

Isolating the ligand as .mol2 file

  • extract the ligand structure from the .pdb file.
  • assign hydrogens.
  • assign all atom (Sybyl/TAFF) and bond types.
  • save it as ligandname.mol2 file.

Running omega

  • run OMEGA, but don't ask me how to do that yet.

Running amsol

  • find more information about amsol on its homepage.
  • mkdir ./amsol2
  • Use file2file.py to get the right formal charge to feed to AMSOL. It is also important to change the name, otherwise the original .mol2 file will be overwritten!

file2file.py -g ligandname.mol2 ./amsol2/someothername.mol2

  • edit ./amsol2/someothername.mol2 :
  • delete all lines prior to @<TRIPOS>MOLECULE
  • change line 2 (molecule name) to something of the format ABCD12345678 (four capital letters followed by eight numbers).
  • line 3 should be natoms nbonds 0 0 0
  • the @<TRIPOS>MOLECULE section must consist of exactly 5 lines (adjust by adding/deleting blanks).
  • remove all sections after the @<TRIPOS>BOND section.
  • delete the blank lines between the ATOM and BOND sections, if there are any.
  • run RunAMSOL3.csh WAIT
  • the output someothername.solv file will contain the following:
AMSOL output
line #1 molname <math>n_{atoms}</math> charge pol_solv ? apol_solv total_solv
other lines charge pol_solv ? apol_solv total_solv
(per_atom)


  • furthermore, there will be someothername.nmol2 file which contains the correct partial charges.

Running mol2db

  • edit someothername.nmol2 so that the @<TRIPOS>MOLECULE section consists of exactly 6 lines.
  • edit the inhier file so that the 'mol2_file', 'db_file' and 'solvation_table' entries are correct.
  • run mol2db inhier
  • add the preamble at the top of the file.
  • gzip the resulting file so that it can be used by DOCK .