I think I'm just getting these or copies or something. We're all set, right? the user can have as many models as he or she wishes, display as many conformations on top of each other as desired, combine As and Bs as desired to make "designer" conformations, etc., etc.
On May 12, 2006, at 10:14 AM, Miguel wrote:
Q: In the general case, can the 'A' atoms from the first run/branch be
paired with the 'B' atoms from the second run/branch?
Is this a valid confomation?
' '
' '
'A'
'A'
' '
' '
'B'
'B'
'B'
' '
' '
My .02:
It seems to me that the most common answer is that, yes, it is or could be a valid conformation; but the question is beside the point... because it is a judgement call that Jmol cannot make.
The person who can decide if the separated A and B runs can be combined legitimately is the user. In fact it may not be possible to determine this without extensive knowledge of the file and the author's intent. This is not a decision, IMO, for Jmol to have to make.
Give the user the power to select these entities, and the ability to see them right off the bat (i.e., render them in the default view, or enable the user to issue a set command that would do so). Then the alternate locations can be handled as desired. It is not necessary to generate a model of some sort for each combination.
Using Miguel's notation above, I think there should be two models that Jmol "sees".
Model 1 is
' '
' '
'A'
'A'
' '
' '
'A'
'A'
'A'
' '
' '
and Model 2 is
' '
' '
'B'
'B'
' '
' '
'B'
'B'
'B'
' '
' '
Could this, the simpler mechanism, be put in place while a more complex combinatorial solution is considered?
Frieda
PS, Regarding the combinatorial solution, perhaps Jmol could generate models up to a certain number that can reasonably be handled well, and if the number of models exceeds that, Jmol can use the non-combinatorial approach... ?
///////////////////////////////////////////
Frieda Reichsman, PhD
Molecules in Motion
Interactive Molecular Structures
///////////////////////////////////////////
