Proteins: Optimizing the positions of the hydrogen atoms

After hydrogen atoms have been added to a protein structure (using ADD-H or by running a PDB file), and all obvious errors corrected, the next step is to optimize their position.  This operation is necessary because the utility that adds hydrogen atoms is very simple, and there are always faults in the positions of the hydrogen atoms.  The positions of the hydrogen atoms can be optimized in a special calculation that holds the rest of the system frozen.  This is done using keywords "NOOPT OPT-H MOZYME GNORM=20 PDBOUT"  plus any keywords generated by the ADD-H operation (typically CHAINS=(text)and START_RES=(text)) These keywords have the following meaning:


Mark all atoms as being frozen


Mark all hydrogen atoms to be optimized


Use the linear-scaling SCF method


Set the geometry optimization criterion to 20


A file in PDB format should be made


The original PDB chain letters should be used


The original PDB residue numbers should be used

An optional. but useful, extra keyword is PLPL allows the SCF calculation to be monitored.

This operation is important, and is absolutely essential if there are any water molecules in the system.  The hydrogen atoms in water molecules are positioned without regard to the rest of the system, i.e., in all water molecules the hydrogen atoms have the same orientation.  Obviously, this is not ideal, and equally obviously the positions of the hydrogen atoms should be optimized before any further work is done.

During the run, the heat of formation will drop, often by thousands of Kcal per mole.  This is expected: the hydrogen atoms were added using a simple procedure, and quite often close contacts are produced.  Optimizing the positions of the hydrogen atoms relieves stresses due to faults in the hydrogenation procedure.  The run should produce a PDB file and an ARC file, as well as the normal output file.

After the positions of the hydrogen atoms have been optimized, check the resulting log, output, and archive files for any problems.  To test if salt bridges should exist, create a salt bridge (using SITE) an re-optimize the positions of the hydrogen atoms.  If the heat of formation decreases, the salt bridge should exist.  Of course, in order to allow heats of formation to be compared, the set of keywords used should be the same, in particular if EPS=78.4 is used in one job it should be used in the other.  This test could be delayed until an unconstrained geometry optimization is done.