Cheri M Turman wrote: > Modeling SO4 explicitly may be more complicated than I thought...not > sure? However, I looked into what I guess are the CHARMM topology > (top_all22_prot_b5.inp) and parameter (par_all22_prot_b5.inp) files that > MODELLER uses by default and there was no entry for SO4. This is > puzzling since SO4 is defined in the retyp.lib file. Has anyone been > successful with changing SO4 single letter PDB entry in the restyp.lib > and having MODELLER use top and parms for it? If so, where are the top > and parms for SO4 located? Maybe I just overlooked them??

SO4 is not exactly 'defined' in the restyp.lib file; as you noticed, its code is '#' (which usually means 'undefined' in Modeller). By default, Modeller will probably use top_heav.lib and par.lib, although these are derived from the CHARMM files you looked at, and thus don't have SO4 parameters or topology.

> Anyway, I was able to get the SO4s to work using 'BLK' notation with '.' > in the .ali file. If I read correctly, this means that there is no > topology or parameters associated with 'BLK' so they are modeled as > rigid bodies based on the template structure coordinates. They appear > fine in the output coordinates so I am happy to proceed with this > process. But, I would still like to know if there are CHARMM top and > parms for SO4 if anyone knows.

That's correct. You will need to add the topology and parameters for SO4 if you want it to be treated as non-rigid (there's a FAQ for this).

Ben Webb, Modeller Caretaker