ADL: must add hydrogens to ligands?

Michael Popov popov at
Thu Feb 10 23:12:14 PST 2005

There is an interesting thing -- The manual recommends to add _all_
hyrogens to the ligand whereas the classic paper about blind docking
using Autodock (Protein Science (2002), 11:1729–1737) uses the
polar-only hydrogen model for ligands. One of the successful
theoretical work with Autodock (Biochemistry 2004, 43, 14521-14528),
and only polar hydrogens again. My own dockings with peptide
ligands with Kollman charges assigned indicate that models with only
polar hydrogens dock better.

But is it due to forcefield optimization for ligands with polar
hydrogens (and there is a mistake in the manual in this case) or the
ligands with ALL hydrogens dock worse just because just have more
rotational degrees of freedom?

It would be great to know the Autodock developer's opinion on this


P.S. The docking examples with 1hvr and 3pdb that comes with Autodock
have only the polar hydrogens assigned for ligands.

Thursday, February 10, 2005, 6:42:15 PM, you wrote:

ybso> Dear all,

ybso> I have a question here. As indicated in the manual, hydrogens must be
ybso> added to all atoms in the ligand to ensure the completeness their
ybso> valences.
ybso> Wondering if this is absolutely necessary. If yes, what kind of software
ybso> packages can i use? Currently i download the ligand pdb files from chemID
ybso> database, and used the java applet they provided to add H. Is this way
ybso> fine?

ybso> Thanks a lot!

ybso> Best Regards,

ybso> yuchen
ybso> ________________________________________________
ybso> --- ADL: AutoDock List  ---
ybso> ---

Best regards,
 Michael                            mailto:popov at

More information about the autodock mailing list