jmol-users — Announcements of new versions, communications about feature requests, etc.

Re: [Jmol-users] How to write one frame as a .mol file?

[Robert H. <ha...@st...>]

If you have separate frames, which I think you do, just make sure you use

select 1.1
write "1.1.mol"

or something similar to that to select only the one frame.

If you have the frame as the current frame, you can use:

select this

or (as long as all the atoms are visible)

select visible


Bob




On Thu, Oct 16, 2025 at 12:37 PM John Keller via Jmol-users <
jmo...@li...> wrote:

> I have a large trajectory file as .xyz and I need to save one frame as a
> .mol file. I swear I have done this before, but now when I try this, either
> nothing is selected, or the whole trajectory is saved. Any suggestions?
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>


-- 
Robert M. Hanson
Professor of Chemistry, Emeritus
St. Olaf College
Northfield, MN
http://www.stolaf.edu/people/hansonr


If nature does not answer first what we want,
it is better to take what answer we get.

-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900

*We stand on the homelands of the Wahpekute Band of the Dakota Nation. We
honor with gratitude the people who have stewarded the land throughout the
generations and their ongoing contributions to this region. We acknowledge
the ongoing injustices that we have committed against the Dakota Nation,
and we wish to interrupt this legacy, beginning with acts of healing and
honest storytelling about this place.*

[Jmol-users] How to write one frame as a .mol file?

[John K. <jwk...@al...>]

I have a large trajectory file as .xyz and I need to save one frame as a
.mol file. I swear I have done this before, but now when I try this, either
nothing is selected, or the whole trajectory is saved. Any suggestions?

Re: [Jmol-users] Changing label size of all atoms

[Francesco P. <chi...@gm...>]

Hi John
your suggested
$delect hydrogen,carbon,oxygen,nitrogen
$ font label 15

did the job, saving me a lot of time

Thanks
Francesco

On Fri, Oct 10, 2025, 11:00 John Keller via Jmol-users <
jmo...@li...> wrote:

> Just zooming the molecule changes the relative size of the labels.
> Try this:
> load $glycerol;
> background white; color bonds gainsboro;
> select carbon,oxygen;
> label "%[atomno]"; set labelalignment center;
> font label 20 bold; color label yellow;
> ---------
> Then zoom it with mouse wheel.
>
> JOhn K.
>
>
> On Thu, Oct 9, 2025 at 11:48 PM Francesco Pietra <chi...@gm...>
> wrote:
>
>> Hello
>> Is any command to change the size of labels for all atoms at once?
>>
>> I have such a request by a referee about a submitted paper, as he wants
>> to see the molecule better. Displacing the labels is not enough
>>
>> Thanks
>>
>> Francesco Pietra
>> _______________________________________________
>> Jmol-users mailing list
>> Jmo...@li...
>> https://lists.sourceforge.net/lists/listinfo/jmol-users
>>
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>

Re: [Jmol-users] Changing label size of all atoms

[lisensky <lis...@be...>]

You can also use

select all;

set fontsize 18;


On 10/10/25 2:47 AM, Francesco Pietra wrote:
> Hello
> Is any command to change the size of labels for all atoms at once?
>
> I have such a request by a referee about a submitted paper, as he 
> wants to see the molecule better. Displacing the labels is not enough
>
> Thanks
>
> Francesco Pietra

-- 
George Lis...@be...
Chemistry Department            608-363-2225
Beloit College
700 College Street
Beloit, WI 53511

Robert H. Solem Emeritus Professor in the Sciences
http://chemistry.beloit.edu/classes/lisensky

Re: [Jmol-users] Changing label size of all atoms

[John K. <jwk...@al...>]

Just zooming the molecule changes the relative size of the labels.
Try this:
load $glycerol;
background white; color bonds gainsboro;
select carbon,oxygen;
label "%[atomno]"; set labelalignment center;
font label 20 bold; color label yellow;
---------
Then zoom it with mouse wheel.

JOhn K.


On Thu, Oct 9, 2025 at 11:48 PM Francesco Pietra <chi...@gm...>
wrote:

> Hello
> Is any command to change the size of labels for all atoms at once?
>
> I have such a request by a referee about a submitted paper, as he wants to
> see the molecule better. Displacing the labels is not enough
>
> Thanks
>
> Francesco Pietra
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>

Re: [Jmol-users] Changing label size of all atoms

[Norwid B. <nb...@ya...>]

On Fri, 10 Oct 2025 09:47:47 +0200
Francesco Pietra <chi...@gm...> wrote:

> Hello
> Is any command to change the size of labels for all atoms at once?

General pattern here is `font label XY monospaced`.  Applied in a MWE which
changes the labels to very large, later to very small:

```
$load $pyridine;
label;
font label 30 monospaced;
font label 10 monospaced;
```


Best regards,

Norwid

[Jmol-users] Changing label size of all atoms

[Francesco P. <chi...@gm...>]

Hello
Is any command to change the size of labels for all atoms at once?

I have such a request by a referee about a submitted paper, as he wants to
see the molecule better. Displacing the labels is not enough

Thanks

Francesco Pietra

Re: [Jmol-users] "Best" way to get the current model set?

[Powell, H. <har...@im...>]

Hi Bob

That is exactly what I need.

Many thanks.

Harry Powell
—
Dr Harry Powell
Structural Bioinformatics Group
Dept. Life Sciences
Imperial College London
har...@im...

Phyre2 Developer
www.imperial.ac.uk/phyre2 <http://www.imperial.ac.uk/phyre2>
Twitter for server news
twitter.com/phyre2server <http://twitter.com/phyre2server>

Re: [Jmol-users] "Best" way to get the current model set?

[Robert H. <ha...@st...>]

I think modelInfo is all you want:

$ x = getProperty("modelInfo")
$ print x.modelCount

*   3*$ print x.models[1]














*{  "_ipt"  :  0  "atomCount"  :  327  "bondCount"  :  337  "chainCount"  :
 1  "file"  :  "https://files.rcsb.org/download/1crn.pdb
<https://files.rcsb.org/download/1crn.pdb>"  "file_model"  :  "1.1"
"groupCount"  :  46  "moleculeCount"  :  1  "name"  :  "1CRN"  "num"  :
 1000001  "polymerCount"  :  1  "title"  :  "1CRN"  "vibrationVectors"  :
 false }*
So I think that is what you need. If for Python via JavaScript, perhaps
using:

info = JSON.parse(Jmol.evaluateVar(jmolApplet0,
"getProperty('modelinfo').format('JSON')"))

would be the most efficient way to get all your needed information.

Bob

Re: [Jmol-users] "Best" way to get the current model set?

[Powell, H. <har...@im...>]

Hi Angel

I think that would probably do the job. I need to declare your "fileControl" variable as a global, then it should work.

I'm slightly surprised there isn't a method within JMolScript that would return the mappings, but if I'm the first person to want it, it's not that surprising!

Many thanks

Harry Powell
—
Dr Harry Powell
Structural Bioinformatics Group
Dept. Life Sciences
Imperial College London
har...@im...

Phyre2 Developer
www.imperial.ac.uk/phyre2 <http://www.imperial.ac.uk/phyre2>
Twitter for server news
twitter.com/phyre2server <http://twitter.com/phyre2server>

Re: [Jmol-users] "Best" way to get the current model set?

[Angel H. <ang...@ua...>]

Hello Harry

If you do the loading from Javascript calling a JmolScript "load" 
command, you could create the control variable in JS storing into it 
each filename and frame number:

var fileControl = {}

j = "1.1";
f = "molecA.pdb";
fileControl[j] = f;
Jmol.script(myJmol, "load " + f);

j = "2.1";
f = "molecB.pdb";
fileControl[j] = f;
Jmol.script(myJmol, "load append " + f);

etc

Or even call those JS from inside JmolScript commands, or create all 
variables in JmolScript, or run them from within JSmol via the Info. 
loadStructCallback

How does it sound?

·
  Dr. Angel Herráez
  Biochemistry and Molecular Biology,
  Dept. of Systems Biology, University of Alcalá
  E-28805 Alcalá de Henares  (Madrid), Spain
  https://biomodel.uah.es/

Re: [Jmol-users] "Best" way to get the current model set?

[Powell, H. <har...@im...>]

Hi

No, not really what I meant. I need to know what the numbers would be for the frame commands and be able to map them to the coordinate file that they correspond to.

Sorry if the following is not very clear! 

None of this is done through the console, it's all controlled by a mixture of Python and JavaScript scripts, so I need to code it up to allow for an arbitrary number of extra models once my base set has been loaded.

In my application, I start off by loading a bunch of PDB files and displaying the contents (using a "load files" command, followed by a number of "select model" commands to render each model according to the type of molecule [protein, ligand, cavity etc). 

My application then creates a bunch of new PDB files according to whatever the user inputs (this runs an external application, so these don't even exist at the time the original load & select commands are run), and I need to be able to display these (from one of my JavaScript modules) - presumably using a "load APPEND" command.

I need to be able to use a "select model" for each of these new molecules in order to render them properly - but I need to be able to identify which model is associated with which (new) PDB file.

So I guess what I need to be able to do is to obtain the number of models that have already been loaded (perhaps using Jmol.getPropertyAsArray(jmolApplet0, "modelInfo"), then the .modelCount value), but it would be most robust if there was a tool where I could get an object that directly maps models to files and contains something like {"1.1": "protein.pdb", "2.1": "cavity.pdb", "3.1": "ligand_pose_1.pdb", "4.1": "ligand_pose_2.pdb"...}.

BTW, I noticed that in the example for getProperty (https://chemapps.stolaf.edu/jmol/docs/#getproperty), the loop uses for (int i = 0; ...) - my JavaScript doesn't allow "int" as a variable declaration.

Harry Powell
—
Dr Harry Powell
Structural Bioinformatics Group
Dept. Life Sciences
Imperial College London
har...@im...

Phyre2 Developer
www.imperial.ac.uk/phyre2 <http://www.imperial.ac.uk/phyre2>
Twitter for server news
twitter.com/phyre2server <http://twitter.com/phyre2server>
 

On 10/09/2025, 22:54, "Robert Hanson via Jmol-users" <jmo...@li...> wrote:

    This email from jmo...@li... originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. If you trust the sender, add them to your safe senders list <https://spam.ic.ac.uk/SpamConsole/Senders.aspx> to disable email stamping for this address. 

    If you just mean 

    1) load a model

    2) load another model and be able to see both

    3) iterate (2), displaying some subset of models


    That's no problem. If that's not what you mean, say more.

    load =1crn
    load APPEND =1rcb
    load append =1apb
    color property modelindex
    frame all
    frame [1 3]
    frame [1 2]
    frame [2 3]

    etc.

Re: [Jmol-users] "Best" way to get the current model set?

[Robert H. <ha...@st...>]

If you just mean

1) load a model

2) load another model and be able to see both

3) iterate (2), displaying some subset of models


That's no problem. If that's not what you mean, say more.

load =1crn
load APPEND =1rcb
load append =1apb
color property modelindex
frame all
frame [1 3]
frame [1 2]
frame [2 3]

etc.

[Jmol-users] "Best" way to get the current model set?

[Powell, H. <har...@im...>]

I was wondering what the "best" way to get the current model set is? I notice that the docs for "show MODEL" say 

	"This command is still in development. The exact form and content of the output is subject to change (and suggestion)."

My problem is that I have a bunch of loaded coordinate files, but at various times (and in different bits of code written in different languages) I want to load more files so that I can view new molecules in the same display. 

Harry Powell
—
Dr Harry Powell
Structural Bioinformatics Group
Dept. Life Sciences
Imperial College London
har...@im...

Phyre2 Developer
www.imperial.ac.uk/phyre2 <http://www.imperial.ac.uk/phyre2>
Twitter for server news
twitter.com/phyre2server <http://twitter.com/phyre2server>

Re: [Jmol-users] When will AI help with Jmol commands? (& Missing Residues vs. AlphaFold)

[<m0...@ya...>]

 Well, this was an interesting AI experiment! Thanks, Camps, for sharing your results! And to Ángel and Bob for diagnosing issues in the AI-generated code.
I implemented ellipsoids to represent missing residues in a FirstGlance update released April, 2020. When I try using Jmol commands for the first time, I usually need help to get the correct syntax. In the case of ellipsoids, Karsten Theis showed me how. So I offered the ellipsoid missing residues question for an AI test because I knew it is challenging. If you'd like to see some nice ellipsoids representing missing residues within protein chains,
https://proteopedia.org/wiki/fgij/fg.htm?mol=6qj4 
The spheres represent missing ends of chains.
I call both the ellipsoids and spheres "empty baskets" as a reminder that they represent missing residues.
During the past year, I wrotehttps://proteopedia.org/w/Missing_residues_and_incomplete_sidechains 
which shows snapshots illustrating how FirstGlance makes you less likely to overlook missing residues than do PyMOL, ChimeraX, iCn3D, or Mol*. The latter four give you no indication of missing sidechains (typically on the folded protein surface), while FirstGlance marks them S- (S minus for missing Sidechain). I suspect that many users of the latter four may spend hours analyzing a structure without realizing there are missing residues and incomplete sidechains (missing surface charges).
In the Proteopedia article I recommend comparing AlphaFold models (which don't have missing residues or incomplete sidechains) to empirical models, especially when looking at shape or electrostatic potential maps. There I show snapshots comparing EPMs between empirical models and AlphaFold models.
FirstGlance has many unique capabilities and conveniences absent in the other four programs mentioned above:https://proteopedia.org/wiki/fgij/whatis.htm#unique 
-Eric

Re: [Jmol-users] Select atoms in common between two models?

[Robert H. <ha...@st...>]

Something like this might get you started:


x = {1.1 and *.CA}.label("%[group]%[resno]:%[chain]!%[model]")
y = {2.1 and *.CA}.label("%[group]%[resno]:%[chain]!%[model]")
z = (x + y).pivot

Now z is a sorted map:

$ print z
{
  "ALA100:C!1.1"  :  60
  "ALA100:C!2.1"  :  60
  "ALA102:A!1.1"  :  60
  "ALA102:A!2.1"  :  60
  "ALA103:B!1.1"  :  60
  "ALA103:B!2.1"  :  60
  "ALA103:C!1.1"  :  60
  "ALA103:C!2.1"  :  60
  "ALA109:B!1.1"  :  60
  "ALA109:B!2.1"  :  60
  "ALA114:A!1.1"  :  60
  "ALA114:A!2.1"  :  60
  "ALA127:B!1.1"  :  60
  "ALA127:B!2.1"  :  60

*  "ALA12:A!1.1"  :  60*  "ALA130:C!1.1"  :  60
  "ALA130:C!2.1"  :  60
...
}


You can then take them two at a time and see if the second one is the same
as the first in front of the "!".
Here we see that only 1.1 has ALA12:A.
But there may also be a count you are looking for. Maybe those have to
match as well?

Bob

Re: [Jmol-users] Select atoms in common between two models?

[Robert H. <ha...@st...>]

I'm not clear on what "same" means here.

- same position?
- same label?
- something else?

What would be a string value that would define "same"?

Re: [Jmol-users] When will AI help with Jmol commands?

[Robert H. <ha...@st...>]

      var length = Math.sqrt(direction.x*direction.x +
direction.y*direction.y + direction.z*direction.z);

there is no "Math.sqrt" -- I think the machine is using JavaScript as a
basis.

Math.min(3,4) does not exist, nor does min(3,4)
It would be [3,4].min

You have to use the ELLIPSOID command, not DRAW.
And the calculation of the ellipsoids does not work like this. You need
three perpendicular axes, but they won't be x, y, and z.

and there are other problems.

Here is a modified script that works. Mind you, other than the actual
format of the commands, the logic is pretty good. Could I have written a
script faster without this flawed template? Maybe not.


[image: image.png]

Re: [Jmol-users] When will AI help with Jmol commands?

[Robert H. <ha...@st...>]

      var direction = {(pos2.x - pos1.x), (pos2.y - pos1.y), (pos2.z -
pos1.z)}

You can't put variable expressions in { xxx, xxx, xxx} points.

      var direction = point((pos2.x - pos1.x), (pos2.y - pos1.y), (pos2.z -
pos1.z));

Likewise for the draw command:

      draw @objectName ellipsoid
        {midX midY midZ}
        {(ellipsoidLength * direction.x) (ellipsoidLength * direction.y)
(ellipsoidLength * direction.z)}
        {ellipsoidWidth 0 0}
        {0 ellipsoidHeight 0};

needs to be (with no linefeeds)

      draw @objectName ellipsoid
        @{point(midX midY midZ)}
         @{point( (ellipsoidLength * direction.x) (ellipsoidLength *
direction.y) (ellipsoidLength * direction.z))}
         @{point( ellipsoidWidth 0 0)}
         @{point( 0 ellipsoidHeight 0)};
and

  for (var i = 1; i < atomCount; i++) {
    var atom1 = sortedAtoms[i-1];
    var atom2 = sortedAtoms[i];

looks a lot like Java or JavaScript, but it is not. When i = 1, i-1 is 0,
which is not the first element, it's the last element.

var allChains = {*}.chains;
print "Processing chains: " + allChains;

Yeiks! That's for all atoms. that needs to be {*}.chain.pivot.keys

and

function createEllipsoidsForChain(chainID) {

  # Get all CA atoms for this chain, sorted by residue number
  var caAtoms = {*.CA and :chainID};
  var atomCount = caAtoms.size;

  if (atomCount < 2) return;

  # Convert to array and sort by residue sequence number
  var sortedAtoms = caAtoms.sort("resno");

sounds great, but there is no command {atoms}.sort("resno") that would
return something like [({5}),({3}),...]


and you can't say this:

  var caAtoms = {*.CA and :chainID};

where chainID is a constant.

etc etc




On Mon, Sep 8, 2025 at 2:27 PM Angel Herraez <ang...@ua...> wrote:

> Astonishing!
>
> However, Jmol throws an error with both scripts (line 32 in the simple
> one) I do not see how that code is bad formed.
> I tried with 2ace and 5nyp
>
>
>
>  Dr. Angel Herráez
>  Biochemistry and Molecular Biology,
>  Dept. of Systems Biology, University of Alcalá
>  E-28805 Alcalá de Henares  (Madrid), Spain
> ------------------------------
> *De:* I. Camps <ic...@gm...>
> *Enviado:* lunes, 8 de septiembre de 2025 20:06
> *Para:* jmo...@li... <jmo...@li...
> >
> *Asunto:* Re: [Jmol-users] When will AI help with Jmol commands?
>
> *ATENCIÓN*: Este correo electrónico se envió desde fuera de la UAH. No
> haga clic en enlaces ni abra archivos adjuntos a menos que reconozca al
> remitente y sepa que el contenido es seguro.
> Hi.
>
> I started playing with a system dedicated to vibe coding called Greta (
> https://greta.questera.ai/home). I asked "Write a Jmol script that make
> ellipsoids between the alpha carbons of the residues within each chain,
> between which consecutive residues are missing", and it returns the
> following:
>
> I've created a comprehensive Jmol script that generates ellipsoids between
> alpha carbons where consecutive residues are missing within each protein
> chain. The script includes:
> Key Features:
>
> 1.**Automatic Gap Detection**: Compares residue numbers to find missing
> sequences
> 2.**Chain-Specific Processing**: Handles multi-chain structures
> independently
> 3.**Visual Ellipsoids**: Creates appropriately sized ellipsoids spanning
> gaps
> 4.**Color Coding**: Yellow (2-3 missing), Orange (4-10 missing), Red (>10
> missing)
> 5.**Semi-transparency**: For better visualization of the underlying
> structure
>
> Three Files Provided:
>
> 1.**missing_residues_ellipsoids.spt** - Full-featured script with detailed
> functions
> 2.**simplified_version.spt** - Streamlined version for basic usage
> 3.**README.md** - Complete documentation and usage instructions
>
> The script will automatically detect all chains in your protein structure,
> identify gaps between consecutive residues, and create appropriately
> colored ellipsoids to visualize these missing regions. Each ellipsoid's
> size is proportional to the distance between the flanking residues and the
> number of missing residues.
>
> The created files are attached here.
>
> []'s,
> Camps
> iv>
>
>
> On Sun, Sep 7, 2025 at 7:19 PM Theis, Karsten W. via Jmol-users <
> jmo...@li...> wrote:
>
> Eric,
>
> We are not that close; ask again in 6 months. I gave Jmol Jane your
> prompt. Then, I had to give it an example script with a for loop. Finally,
> it came up with a script (which I did not check). Anyway, I think it is
> possible to write a script like that. You also need to figure out the
> chains and apply the idea separately to the chains. AI transcript link is
> here: https://chatgpt.com/share/68bdffed-376c-800a-924a-180f1afc4dba
>
> Karsten
> ------------------------------
> *From:* m0lviz--- via Jmol-users <jmo...@li...>
> *Sent:* Sunday, September 7, 2025 17:27
> *To:* Robert Hanson via Jmol-users <jmo...@li...>
> *Cc:* m0...@ya... <m0...@ya...>
> *Subject:* [Jmol-users] When will AI help with Jmol commands?
>
>
> Caution External Email: This email originated outside of WSU. Do not
> click links, open attachments, or respond if it appears to be suspicious.
> Are we close to it being feasible to train an AI system to understand Jmol
> command language well enough that it can generate working code to
> accomplish a rendering phrased in English?
>
> For example, "What are the the Jmol commands to make ellipsoids between
> the alpha carbons of the residues within each chain, between which
> consecutive residues are missing?"
>
> Will it be possible for the AI system to learn about Jmol command
> capabilities that are implemented but not documented? (From comments Bob
> Hanson has made, my impression is that there are lots of those.)
>
> -Eric
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>


-- 
Robert M. Hanson
Professor of Chemistry, Emeritus
St. Olaf College
Northfield, MN
http://www.stolaf.edu/people/hansonr


If nature does not answer first what we want,
it is better to take what answer we get.

-- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900

*We stand on the homelands of the Wahpekute Band of the Dakota Nation. We
honor with gratitude the people who have stewarded the land throughout the
generations and their ongoing contributions to this region. We acknowledge
the ongoing injustices that we have committed against the Dakota Nation,
and we wish to interrupt this legacy, beginning with acts of healing and
honest storytelling about this place.*

Re: [Jmol-users] When will AI help with Jmol commands?

[Angel H. <ang...@ua...>]

Astonishing!

However, Jmol throws an error with both scripts (line 32 in the simple one) I do not see how that code is bad formed.
I tried with 2ace and 5nyp



 Dr. Angel Herráez
 Biochemistry and Molecular Biology,
 Dept. of Systems Biology, University of Alcalá
 E-28805 Alcalá de Henares  (Madrid), Spain
________________________________
De: I. Camps <ic...@gm...>
Enviado: lunes, 8 de septiembre de 2025 20:06
Para: jmo...@li... <jmo...@li...>
Asunto: Re: [Jmol-users] When will AI help with Jmol commands?

ATENCIÓN: Este correo electrónico se envió desde fuera de la UAH. No haga clic en enlaces ni abra archivos adjuntos a menos que reconozca al remitente y sepa que el contenido es seguro.
Hi.

I started playing with a system dedicated to vibe coding called Greta (https://greta.questera.ai/home). I asked "Write a Jmol script that make ellipsoids between the alpha carbons of the residues within each chain, between which consecutive residues are missing", and it returns the following:

I've created a comprehensive Jmol script that generates ellipsoids between alpha carbons where consecutive residues are missing within each protein chain. The script includes:
Key Features:

1.**Automatic Gap Detection**: Compares residue numbers to find missing sequences
2.**Chain-Specific Processing**: Handles multi-chain structures independently
3.**Visual Ellipsoids**: Creates appropriately sized ellipsoids spanning gaps
4.**Color Coding**: Yellow (2-3 missing), Orange (4-10 missing), Red (>10 missing)
5.**Semi-transparency**: For better visualization of the underlying structure

Three Files Provided:

1.**missing_residues_ellipsoids.spt** - Full-featured script with detailed functions
2.**simplified_version.spt** - Streamlined version for basic usage
3.**README.md** - Complete documentation and usage instructions

The script will automatically detect all chains in your protein structure, identify gaps between consecutive residues, and create appropriately colored ellipsoids to visualize these missing regions. Each ellipsoid's size is proportional to the distance between the flanking residues and the number of missing residues.

The created files are attached here.

[]'s,
Camps
iv>


On Sun, Sep 7, 2025 at 7:19 PM Theis, Karsten W. via Jmol-users <jmo...@li...<mailto:jmo...@li...>> wrote:
Eric,

We are not that close; ask again in 6 months. I gave Jmol Jane your prompt. Then, I had to give it an example script with a for loop. Finally, it came up with a script (which I did not check). Anyway, I think it is possible to write a script like that. You also need to figure out the chains and apply the idea separately to the chains. AI transcript link is here: https://chatgpt.com/share/68bdffed-376c-800a-924a-180f1afc4dba

Karsten
________________________________
From: m0lviz--- via Jmol-users <jmo...@li...<mailto:jmo...@li...>>
Sent: Sunday, September 7, 2025 17:27
To: Robert Hanson via Jmol-users <jmo...@li...<mailto:jmo...@li...>>
Cc: m0...@ya...<mailto:m0...@ya...> <m0...@ya...<mailto:m0...@ya...>>
Subject: [Jmol-users] When will AI help with Jmol commands?


Caution External Email: This email originated outside of WSU. Do not click links, open attachments, or respond if it appears to be suspicious.

Are we close to it being feasible to train an AI system to understand Jmol command language well enough that it can generate working code to accomplish a rendering phrased in English?

For example, "What are the the Jmol commands to make ellipsoids between the alpha carbons of the residues within each chain, between which consecutive residues are missing?"

Will it be possible for the AI system to learn about Jmol command capabilities that are implemented but not documented? (From comments Bob Hanson has made, my impression is that there are lots of those.)

-Eric
_______________________________________________
Jmol-users mailing list
Jmo...@li...<mailto:Jmo...@li...>
https://lists.sourceforge.net/lists/listinfo/jmol-users

Re: [Jmol-users] When will AI help with Jmol commands?

[Robert H. <ha...@st...>]

On Sun, Sep 7, 2025 at 4:27 PM m0lviz--- via Jmol-users <
jmo...@li...> wrote:

>
> Will it be possible for the AI system to learn about Jmol command
> capabilities that are implemented but not documented? (From comments Bob
> Hanson has made, my impression is that there are lots of those.)
>
>
FYI, my undergraduate interns in January brought the documentation up to
date; there is more to add since then, but nothing that would interest you,
I think, Eric. Just this feature:

new feature: BCIF, mmCIF, and MMTF file readers auto-detect carbohydrate
residues

All the 36 other new features since January relate to specialized functions
relating mostly to crystallography. But that does remind me that I should
add those. I'll see what I can do.

As for finding undocumented features, that would certainly be a trick. But
what would be the training set? The actual Java code?

I asked Google AI:* how does the matrix.array() function work in Jmol*

and it responded, referencing the Jmol/JSmol Interactive Documentation:

In Jmol scripting, the
matrix.array() function returns the contents of a matrix variable as a
one-dimensional array. This is an important distinction, because Jmol
treats matrices differently from arrays, especially when dealing with
specific dimensions like 3x3 or 4x4.
How it works

   - *Variable types*: In Jmol's JavaScript-like environment, variables can
   be defined as matrices or arrays. A matrix is specifically for numerical
   values and is defined with double brackets, like
   [[1,2,3],[4,5,6],[7,8,9]]. An array, on the other hand, is a more
   general list that can contain mixed variable types and is created using the
   array() function or single brackets, ["a", "b", 3].
   - *Dimensionality*: A matrix is a two-dimensional object. The
   matrix.array() function flattens this two-dimensional structure into a
   single list of numbers. The elements are extracted row by row, from left to
   right.
   - *Practical application*: This function is useful when you need to
   perform an operation on the individual numerical components of a
   transformation matrix, but the operation requires a standard
   one-dimensional array as input. For example, if you have a 4x4 matrix
   representing a translation and rotation, matrix.array() will extract the
   16 numerical values into a single array.


   -

Example
Here is a simplified example of how matrix.array() would be used in a Jmol
script:

# Create a 3x3 matrix
var myMatrix = [[1, 2, 3], [4, 5, 6], [7, 8, 9]];

# Use matrix.array() to convert it to a 1D array
var myFlatArray = myMatrix.array();

# The result is a single array of 9 elements
# The print command shows the contents of the new array
print myFlatArray;

# This script would output:
# [1, 2, 3, 4, 5, 6, 7, 8, 9]


Only problem is that that's not true. matrix.array() returns a 3x3 or 4x4
array of arrays, not a simple array. The documentation *actually says:*

*x = y.arrayForces y into an array format. If y is already an array just
returns that; otherwise returns an array created from y. Variables of type
matrix3f and matrix4f are converted to standard 3x3 or 4x4 arrays; others
are returned as [ xxx ]. This function can be used to force an array where
a string or other single value might be returned. In this way it is similar
to y.all, but whereas y.all only works on atom properties, y.array works on
any type of value.*



alas.



Bob

Re: [Jmol-users] When will AI help with Jmol commands?

[I. C. <ic...@gm...>]

Hi.

I started playing with a system dedicated to vibe coding called Greta (
https://greta.questera.ai/home). I asked "Write a Jmol script that make
ellipsoids between the alpha carbons of the residues within each chain,
between which consecutive residues are missing", and it returns the
following:

I've created a comprehensive Jmol script that generates ellipsoids between
alpha carbons where consecutive residues are missing within each protein
chain. The script includes:
Key Features:

1.**Automatic Gap Detection**: Compares residue numbers to find missing
sequences
2.**Chain-Specific Processing**: Handles multi-chain structures
independently
3.**Visual Ellipsoids**: Creates appropriately sized ellipsoids spanning
gaps
4.**Color Coding**: Yellow (2-3 missing), Orange (4-10 missing), Red (>10
missing)
5.**Semi-transparency**: For better visualization of the underlying
structure

Three Files Provided:

1.**missing_residues_ellipsoids.spt** - Full-featured script with detailed
functions
2.**simplified_version.spt** - Streamlined version for basic usage
3.**README.md** - Complete documentation and usage instructions

The script will automatically detect all chains in your protein structure,
identify gaps between consecutive residues, and create appropriately
colored ellipsoids to visualize these missing regions. Each ellipsoid's
size is proportional to the distance between the flanking residues and the
number of missing residues.

The created files are attached here.

[]'s,
Camps
iv>


On Sun, Sep 7, 2025 at 7:19 PM Theis, Karsten W. via Jmol-users <
jmo...@li...> wrote:

> Eric,
>
> We are not that close; ask again in 6 months. I gave Jmol Jane your
> prompt. Then, I had to give it an example script with a for loop. Finally,
> it came up with a script (which I did not check). Anyway, I think it is
> possible to write a script like that. You also need to figure out the
> chains and apply the idea separately to the chains. AI transcript link is
> here: https://chatgpt.com/share/68bdffed-376c-800a-924a-180f1afc4dba
>
> Karsten
> ------------------------------
> *From:* m0lviz--- via Jmol-users <jmo...@li...>
> *Sent:* Sunday, September 7, 2025 17:27
> *To:* Robert Hanson via Jmol-users <jmo...@li...>
> *Cc:* m0...@ya... <m0...@ya...>
> *Subject:* [Jmol-users] When will AI help with Jmol commands?
>
>
> Caution External Email: This email originated outside of WSU. Do not
> click links, open attachments, or respond if it appears to be suspicious.
> Are we close to it being feasible to train an AI system to understand Jmol
> command language well enough that it can generate working code to
> accomplish a rendering phrased in English?
>
> For example, "What are the the Jmol commands to make ellipsoids between
> the alpha carbons of the residues within each chain, between which
> consecutive residues are missing?"
>
> Will it be possible for the AI system to learn about Jmol command
> capabilities that are implemented but not documented? (From comments Bob
> Hanson has made, my impression is that there are lots of those.)
>
> -Eric
> _______________________________________________
> Jmol-users mailing list
> Jmo...@li...
> https://lists.sourceforge.net/lists/listinfo/jmol-users
>

Re: [Jmol-users] When will AI help with Jmol commands?

[Theis, K. W. <kt...@we...>]

Eric,

We are not that close; ask again in 6 months. I gave Jmol Jane your prompt. Then, I had to give it an example script with a for loop. Finally, it came up with a script (which I did not check). Anyway, I think it is possible to write a script like that. You also need to figure out the chains and apply the idea separately to the chains. AI transcript link is here: https://chatgpt.com/share/68bdffed-376c-800a-924a-180f1afc4dba

Karsten
________________________________
From: m0lviz--- via Jmol-users <jmo...@li...>
Sent: Sunday, September 7, 2025 17:27
To: Robert Hanson via Jmol-users <jmo...@li...>
Cc: m0...@ya... <m0...@ya...>
Subject: [Jmol-users] When will AI help with Jmol commands?


Caution External Email: This email originated outside of WSU. Do not click links, open attachments, or respond if it appears to be suspicious.

Are we close to it being feasible to train an AI system to understand Jmol command language well enough that it can generate working code to accomplish a rendering phrased in English?

For example, "What are the the Jmol commands to make ellipsoids between the alpha carbons of the residues within each chain, between which consecutive residues are missing?"

Will it be possible for the AI system to learn about Jmol command capabilities that are implemented but not documented? (From comments Bob Hanson has made, my impression is that there are lots of those.)

-Eric

[Jmol-users] When will AI help with Jmol commands?

[<m0...@ya...>]

Are we close to it being feasible to train an AI system to understand Jmol command language well enough that it can generate working code to accomplish a rendering phrased in English?
For example, "What are the the Jmol commands to make ellipsoids between the alpha carbons of the residues within each chain, between which consecutive residues are missing?"
Will it be possible for the AI system to learn about Jmol command capabilities that are implemented but not documented? (From comments Bob Hanson has made, my impression is that there are lots of those.)
-Eric

[Jmol-users] Select atoms in common between two models?

[<m0...@ya...>]

9eab "model B", and 9ead "model D", are two cryo-EM models of a virus capsid in different conformations. Model B has 345,300 atoms in chains A,B,C while model D has 164,940 in those same chains because of poorer resolution.
How do I select the atoms in common between the two models?
$ load =9eab filter "!_H;biomolecule 1"$ load append =9ead filter "!_H;biomolecule 1"$ models all$ select 1.1 and (:A,:B,:C)345300 atoms selected$ select 2.1 and (:A,:B,:C)164940 atoms selected
$ select 1.1 or 2.1526920 atoms selected$ select 1.1 and 2.10 atoms selected # obviously
$ select gly1:B and 1.1240 atoms selected$ select gly1:B and 2.1240 atoms selected
Thus, gly1:B is present in both models.
How do I select all atoms present in both models, ignoring their model numbers?
Thanks, -Eric