[Jmol-developers] [ jmol-Feature Requests-964842 ] support internal coordinates

[SourceForge.net <no...@so...>]

Feature Requests item #964842, was opened at 2004-06-02 03:37
Message generated for change (Settings changed) made by hansonr
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=379136&aid=964842&group_id=23629

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
>Status: Closed
Priority: 1
Submitted By: Miguel (migueljmol)
Assigned to: Bob Hanson (hansonr)
Summary: support internal coordinates

Initial Comment:
Need to support internal coordinates for MOPAC (&
other?) files. 

While I am somewhat reluctant to say this, I suspect
that this may be better off done in the ModelReader
code instead of in the core. 

It is not like fractional coordinates, where we need to
have support in the core. 

And this issue of 'artificial atoms' to get started is
a problem. 

It seems to me that we should build a small amount of
support into the ModelReader and/or Model classes. Then
MOPAC can use it. 

The other thing that concerns me is that the
ModelAdapter classes currently do not make any
references to javax.vecmath ... it is supposed to be a
simple implementation. 

So, who knows ... maybe it should be in the core. 

---------------------------- Original Message
---------------------------- Subject: Re:
[Jmol-developers] internal coordinates?
From:    "Bob Hanson" <ha...@st...>
Date:    Tue, June 1, 2004 21:08
To:      jmo...@li...
--------------------------------------------------------------------------

Miguel, all you do is read these data in and
immediately convert to cartesian  coordinates. It's set
up so that you stream the data in. The first three
points  are unique, because:

a) The first point (#1) is automatically just at
position 0,0,0.
b) The second point (#2) is a given distance from that
point, #NA=1 along X c) The third point (#3) is a given
distance from atom #NA=2 with a given angle  in the xy
plane for NA-NB-#3.
d) All additional points (#X) are referenced to
previous points by indicating a  distance from a
previous point (NA-#X), an angle back to two points
(NA-NB-#X),  and a torsional angle back to three
previous points (NA-NB-NC-#X)

see
http://www.cmbi.kun.nl/tutorials/cheminf/mopac/intc.html
for a nice description of a "z-matrix"

Bob


Miguel wrote:

>>internal coordinates SHOULD be very easy to use,
particularly if they are handed
>>to you and you don't have to generate them yourself.
They are amazingly concise.
> 
> 
> We are already handling fractional coordinates, so
adding another coordinate system will not be difficult
... IF ...
> 
> 
>>The only complication I can imagine is that they
sometimes involve "phantom" or
>>"temporary" atoms for starting points.
>>
>>For example, in this example we have a first atom at
0,0,0, then two phantom
>>atoms on roughly perpendicular axes. They define the
starting point for the next
>>dihedral, and you are off and running. Mostly QM
programs start with these so as
>>to define positions when symmetry is an issue.
>>
>>   ATOM    CHEMICAL     BOND LENGTH    BOND ANGLE   
TWIST ANGLE
>>  NUMBER    SYMBOL      (ANGSTROMS)    (DEGREES)    
(DEGREES)
>>    (I)                   NA:I          NB:NA:I     
 NC:NB:NA:I       NA
>>  NB
>>    NC
>>     1       C(  1)        0.000000      0.000000   
   0.000000
>>
>>     2      XX(   )        4.117293      0.000000   
   0.000000        1
>>
>>     3      XX(   )        2.303156     90.000000   
   0.000000        2
>>   1
>>
>>     4       C(  2)        1.351924  * 133.488008  *
 180.000000  *     1
>>   2
> 
> 
> ... IF ... there are three floats per atom.
> 
> Q: Is it the case that there are only 3 numbers per
atom? bondLength, BondAngle, and TorsonAngle?
> 
> Q: What are NA, NB, and NC?
> 
> 
> 
> Miguel
> 

---------------------------- Original Message
---------------------------- Subject: Re:
[Jmol-developers] internal coordinates?
From:    "Egon Willighagen" <eg...@sc...>
Date:    Tue, June 1, 2004 21:00
To:      jmo...@li...
Cc:      "Miguel" <mi...@jm...>
--------------------------------------------------------------------------

On Tuesday 01 June 2004 20:48, Miguel wrote:
> > internal coordinates SHOULD be very easy to use,
particularly if they are handed
> > to you and you don't have to generate them
yourself. They are amazingly concise.
>
> We are already handling fractional coordinates, so
adding another coordinate system will not be difficult
... IF ...
>
> ... IF ... there are three floats per atom.
>
> Q: Is it the case that there are only 3 numbers per
atom? bondLength, BondAngle, and TorsonAngle?

The system works differently...

The first atom normally has nothing
The second has a distance (to the first)
The third has a distance (to the second) and an angle
(first -> second ->  third)
The fourth has a distance (to the third) an angle
(second -> third -> fourth)  and an torsion (first ->
sec -> third -> fourth)

This is called the Z-matrix...

> Q: What are NA, NB, and NC?

That's file format specific... dunno... not Z-Matrix
stuff ...

Egon





----------------------------------------------------------------------

Comment By: Bob Hanson (hansonr)
Date: 2006-08-22 17:39

Message:
Logged In: YES 
user_id=1082841

Jmol 11 will not read z-matrix coordinates 

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=379136&aid=964842&group_id=23629