Bob Hanson wrote:
Is this a "feature" or a "bug"? Should "within" restrict itself to a
given model, or is it interesting or important that one be able to
select "within" a distance across multiple model sets?
it is an important *feature* that one is able to use within across
multiple model sets.
right, that's my point. It seems unnatural to me as well. That's why I
was surprised and initially thought it was a bug. Personally I can't
imagine it, either, unless....
...I suppose some day we will be loading two molecules and maneuvering
them independently (like a drug and an active site). In that case they
will (probably) be in two different frames because they will have been
loaded from two different files, and the interesting thing might be to
highlight what atoms in one molecule are near atoms in another.
this case is called "molecule" in RasMol2.7. and I don't want to load
them as two separate PDB-MODELs in Jmol "animation frames". They could
be addressed for maneuvering independently but should be in the same
MODEL allowing measurements between molecules.
So maybe someday there will be a flag that can be set -- only within a
model, only between models -- still I find it hard to believe one
would ever want the current "within or between models".
I have a feeling that the logic would get totally unrulely if we tried
to restrict "within" to a particular model. Because this has to result
in a linear set of bits.
This is easy
select within(4.2, atomno=3 AND */1) AND */1
Bob
Nicolas Vervelle wrote:
Angel Herraez wrote:
Diificult to distinguish, since most atoms nr. 3 are close to each
other.
I have tested with another file and yes, it is selecting atoms in
any model that are close to specified atom in any model (i.e, not
only within the same model).
Test attached PDB file using atomno=22 as the reference
From my totally non-expert point of view, this behavior seems very
strange and unnatural.
(since different models could be totally unrelated to each other, or
related in different ways)
I don't see why someone would want to select atoms at coordinates
near at atom in an other model.
The PDB allows to spread a *single* *molecule* over *different* *models*
to overcome the 99999 atomno restiction
"
MODEL
...
* If a collection contains more than 99,999 total atoms, then more than one
entry must be made. In such a case the collection is divided between models
(between an ENDMDL and the following MODEL record) and the model
numbering is
sequential throughout such a set of entries.
"
(this is, if the models had been aligned by the user - PDB - in some
special way before) but it does not harm because you can restrict your
selection to only one model if intended and JavaScript may generate your
code, if you need iteration through all models.
In this light, we shouldn't be more papist than PB and allow altLoc
addressing by ignoring the theoretical combinatorial nature of *altLoc*
generated models and generate them *linear*, using all *;D (RasMol ";")
to generate a single all *%D (Jmol "%") altLoc conformation D containing
model.
Regards, Jan
-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Jmol-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-users
