Looks like you have already gotten several good suggestions, but I also
wrote a jiffy program for doing this that does not require the two PDB
files to have the same atom names:
http://bl831.als.lbl.gov/~jamesh/pickup/origins.com
Which you run like this:
origins.com right_origin.pdb wrong_origin.pdb P212121 correlate nochains
The bottom of the script file contains a list of allowed origin shifts,
which are each applied in turn and the resulting symmetry-expanded atom
constellations compared. If you use the word "correlate" on the command
line the atoms will be converted to an electron density map using sfall
and the correlation coefficient used as the "match" score. By default,
the program breaks up the PDBs into their "chains" (segids) and searches
each one separately. You can turn this off by using the word "nochains"
on the command line.
I think "emma" should give you similar results, but the algorithms are
certainly different.
-James Holton
MAD Scientist
Francois Berenger wrote:
Hello,
1) In the case I know the true structure (I am verifying I use Phaser
correctly), is there a program to do so?
Some other questions, if I am to write this program by myself:
2) is there a list somewhere of the translation ambiguities for
each spacegroup?
For example, in P1 it would say me any translation on any axis
is fine.
3) is there a list of permissible origins for each space group?
For example, in P212121 it would say me there are 8 possible choices
and list them for me.
I already know of symop.lib, but I don't think it has some of the
information I am looking for.
I also know the csymmatch example program of the clipper library, it
does part of what I intend to do.
Thanks a lot,
Francois.
