Hi YoungJin,
I was now modifying AMPPNP ('ANP' in its library name). Although it came from coot/ccp4 library, phenix does not recognize so I had to generate .cif file from the pdb. As it turned out AMPPNP is hydrolyzed in the crystal having AMPPN (terminal Phosphate group needs to be removed), I made from coot (simply deleted atoms) and made another .cif file. One question what I have now is while I am adding H atoms into a whole molecule, how I can add H atoms onto terminal AMPPN (N atom). Unlike ATP case, it is very certain that terminal N atom should have 2 H's. I did Molprobity as in the general case to add H atoms, but it didn't make it.
I hope Nigel will comment on this (he is the author of phenix.ready_set - a program that is designed to be smart about adding H atoms to everything and creating corresponding CIF files).
Another questions is how I can remove H atoms from the pdb file.
I'm wondering about why you need to do this? Anyway, here is how: phenix.reduce -trim model.pdb > model_no_h.pdb or phenix.pdbtools model.pdb xxx.cif remove="element H" or equivalently: phenix.pdbtools model.pdb xxx.cif keep="not element H" More about phenix.pdbtools: http://www.phenix-online.org/documentation/pdbtools.htm Pavel.