Hi Cedric, The "match" command fits whatever atoms you specify: http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/match.html> I should mention that MatchMaker (or "matchmaker" command) generally does pretty well even with low sequence identity because it also uses secondary structure assignments: the score favors aligning helix with helix, strand with strand. However, if you need it, "match" can be used in any situation, including molecules other than proteins and nucleic acids. The tradeoff is that it is more work for the user, since you have to tell it which atoms to use. Discussion of the different ways to superimpose structures in Chimera: <http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/superposition.html> I hope this helps, Elaine On Apr 17, 2009, at 8:26 AM, Cédric Govaerts wrote:
Hi Elaine,
So of course I have a new question:
Matchmaker will take to proteins, align the sequences and then align the structures.
But what I need to do is to align the structure according to the pairs of atoms I have selected.
For example this could be particularly useful for :
1/ Superimpose two structures that are different but have a few common catalytic residues. They would be two different for Match maker to find the correct alignment, however, the user (that's me) would know which residues should be used.
2/ I tried to superimpose two isoforms of the same proteins using ions or ligand as superimposing element. I wanted to see what was the conformational change *relative* to the bound metals.
I believe that creating an alignment for MultiAlign would not work for either.
Is there a way I can superimpose two structures after explicitely choosing the pairs (minimum) and the Chimera would make a lest-square regression based on those ?
Thanks again
Cedric