In article <3332E8DA.167E at MolecularSimulationsInc.com>,
bobf at MolecularSimulationsInc.com says...
>>Gregory M. F. Watson wrote:
>>>> I have
>> the pdb coordinate file (1RSC) for RubisCO. This file includes the
>> coordinates for a large (L) and a small (S) subunit. Various
>> transformation matrices are included in the pdb ...
>> My understanding is that Quanta can
>> handle these transformations and generate the desired multimeric
>> structures ...
>>One reason you can't find anything in the QUANTA documentation about
>this is because QUANTA can't do it directly unless the other subunit can
>be generated by a crystallographic symmetry operation (based on my own
>experience, I'd say this is very seldom the case with PDB files; they're
>almost always related by noncrystallographic symmetry). You also have
>to have the crystal modelling option, which is not part of the basic
>QUANTA package.
>>However, it's a common thing to want to do, so I wrote a little CHARMm
>script a while back to handle it. You set a few CHARMm variables to
>tell it what the MATRIX records say, and then stream the script to
>transform the coordinates. It doesn't actually generate a copy, but you
>can write out the new coordinates and then read both molecules into
>QUANTA to make the dimer (or more-mer, but then you need to run the
>script once for each transformation matrix).
I'm more a pure CHARMM person than a Quanta user so I don't
know how applicable the following is but you can make copies
as follows (at least in standalone academic CHARMM 23);
1) set up the initial coordinate set (eg. a segment called SEG1)
2) then prepare for as many copies as you need as follows;
GENERATE SEG2 SETUP DUPLICATE SEG1
GENERATE SEG3 SETUP DULPICATE SEG1
GENERATE SEGn...
3) then read in the first set of coordinates (eg. from PDB file)
4) then copy the coordinates from SEG1 to the others as follows;
COOR DUPLICATE SELE ATOM SEG1 * * END SELE ATOM SEG2 * * END
COOR DUPLICATE SELE ATOM SEG1 * * END SELE ATOM SEG3 * * END
etc.
This will give you several superimposed segments.
5) then shift the duplicate segments as required;
COOR TRANS SELE SEGID SEG2 END XDIR x1 YDIR y1 ZDIR z1
COOR TRANS SELE SEGID SEG3 END XDIR x2 YDIR y2 ZDIR z2
etc.
where x1, y1, z1 etc. are the distances to be shifted along
the appropriate axes.
If you like you can then join all the segments into one big one.
Then save the whole mess to one CRD and PSF file.
The main problem is not over running CHARMM's atom limit. In
"big" charmm this is >60,000. I've made a 30 Angstrom radius
sphere of water for immersing proteins in this manner.
Maybe some food for thought...
Bernard
Bernard Murray, Ph.D.
bernard at elsie.nci.nih.gov (National Cancer Institute, NIH, Bethesda MD, USA)
