On 5/23/14, 10:59 AM, Alexander Parella wrote: > I have been attempting to use Modeller to refine the loops of some > generated protein structures. When performing loop refining, Modeller > seems to remove all hydrogen atoms from the structure. I have not > been able to find any way to change this behavior. Are there any such > options? That is, to stop Modeller from removing the Hydrogen atoms > from the structures it refines?
Loop refinement uses a statistical potential, and that doesn't work on hydrogen atoms, since very few PDB structures contain hydrogens. So the hydrogens would only be constrained by stereochemistry. We typically don't build models in the first place containing hydrogens either, because this roughly doubles the number of atoms in the system and thus roughly quadruples the run time, for little benefit. That said, if you really want loop models containing hydrogens, you can subclass loopmodel in the same way that allhmodel subclasses automodel (check out the definition of the allhmodel class in your copy of Modeller).
Ben Webb, Modeller Caretaker