Macromolecular Modeling Blog ™

UGENE. Voted for others

Just the other day I blogged about Jmol in Bioclipse. This has great potential and we are thinking on a Protopedia plugin to show how documentation and running scripts can be performed in Bioclipse:

http://chem-bla-ics.blogspot.com/2009/03/bioclipse-powerful-jmol-application.html

You can leave comments there on what you find missing in Jmol, which, BTW, also has MM dynamics runs functionality.

Warren and I have talked on several occasions on exchanging ‘views’ between PyMOL and Jmol, but we never found the time to come up with an open standard for doing that.

An open standard for “views” or “sessions” between the different software would be very cool, especially since each is good at a specific task. Imagine you can render your session in PyMOL and embed it in a Jmol applet…

I use PyMol because it’s featureful, easy, popular, has a great community & wiki, and open source under a “normal” license.

I use PyMOL – nice, simple, not ideal for ligand-receptor visualization, I’m using Sybyl for that purpose;

For molecular dynamics I use Chimera and sometimes VMD.

What about PMV (python molecular viewer)? It gives users endless possibilities. All the other programs are pretty rigid, and demand you conform to the prescribed conventions. I think PMV allows you to make the most advanced graphics without the use of another program.

VMD is still the tool I reach for first when I want to visualize something. Probably because it’s what I learned first and I’m most proficient with. I do prefer PyMOL in different scenarios such as movie making. Jmol is also excellent since it’s the most web friendly.

The free viewer from MolSoft has been my favorite for many years.

Pymol is flexible, comes with great support, great wiki, really powerful features including modeling features (if you provide a small amount of support to the developer — which everyone should do if they can). Easy to make movies, display MD trajectories, etc. No experience with other packages, but no interested in trying them since pymol does all I need and a lot more.

For myChimera without a doubt is the easiest one I have tried to integrate into my general chemistry/organic classes.

* Very complete package. (critical for students)
* Excellent, comprehensive help files. (critical for students)
* Logical layout of menus/commands. (critical for students)
* Commands make sense to normal instructors/professors, and students who lack the programming background and don’t have time to pick all that up.
* Great integration with tools like Delphi.
* Absolutely top notch, responsive group of developers who listen to you and work hard to help.

While I have worked a little with some of the other apps. (e.g. PyMol, VMD), nothing I have touched comes close to this for the majority of what I do.

I have been using Chimera from the moment I discovered it, years and years ago. The most interesting thing with Chimera is that when my interests have evolved from the X-ray crystallography field to include other structural fields, namely Molecular Dynamics and Electron Microscopy, I have always found Chimera coping with my new needs. Better: it has put before my eyes simple but powerfool tools to perform analyses useful in these fields that I even didn’t know about.

Chimera is indeed an intuitive graphic interface, but it is also python-scriptable, which is a powerful point.

What could I say about their developpers? Whenever I have requested a feature or encountered a bug, they have answered promptly and effectively. They are nice and efficient people, a jewel nowadays.

There are of course some things yet to be improved, but they are not really limiting and I’m sure I will find them improved some time soon.

Chimera is my choice because:

1) In my opinion it is the easiest to use.

2) It has a lot of functionality. I use it all the time to:
- Compare crystal structures to electron microscopy maps (docking).
- Produce publication quality figures.
- Align and compare similar atomic structures.
- Measure distances & angles between residues in atomic structures.

3)It has a very responsive user group where you can get answers in a few hours or even immediately in many cases.

4) Intallation is pain free in Windows, Linux or Mac systems.

5) You can write scripts into a conveninent common use language (Python).

I use chimera for the following reasons:
1) ability to integrate and manage heterogeneous datasets such as multiscale EM.
2) quick response on problems, such as s/w bugs or operator errors.
3) accurate compliance to standards (e.g., VRML).
4) extensible design, multifaceted.
5) excellent graphics

For easy viewing and education I typically use YASARA. It is insanely fast and easy to use to make pictures. The movie option is easy to make teaching content.
For my research I use Coot and unfortunately VI is also needed frequently for looking at PDB files.

Wow, VI, now that’s extreme! Thanks for your comments everyone, I gotta say I’ve learned about some new features from software I thought I knew well.

I routinely use 3 different viewers for technical work, not sure I can pick one as a favorite.
UCSF Chimera: viewing and aligning large complexes and displaying/fitting electron microscopy maps. (multiscale model, matchmaker, volume viewer and volume tools).
Pymol: beautifully rendered pictures, powerful command language and scripting.
Coot: viewing/manipulating models with X-ray data.

Pymol, beautiful pictures with MOE files

Even if Jmol isn’t as polished, feature-rich, or easy to use as the other offerings, it is one of my favorite viewers because it is available as unrestricted open-source code.

This means its usefulness and availability to you and to others is guaranteed into perpetuity, even if the current developers move on, are de-funded, or experience a change of heart.

In practice, nobody can ever take Jmol away from you, or tell you want you can or can’t do with it, or prevent you from sharing it with others who could benefit from it as well.

Some folks were burnt by reliance upon closed-source viewers of the past (Chime, MidasPlus, WebLabViewer, etc.) and now know better than to rely upon their modern equivalents (ICM-Browser, Chimera, DSVisualizer, SwissPDBViewer, etc.)

But those who don’t know of or haven’t learned from others’ mistakes of the past may eventually be doomed to repeat them.

So if unrestricted open-source code matters to you, then you’ve got to favor Jmol, or perhaps that other open-source viewer… Py…
something.

PyMOL is my first choice for routine visualization, getting images for publications, and for making movies for presentations.

Despite having several years of experience with PyMol, I still need to look up how to do things from examples (http://www.chem.ucsb.edu/~molvisual/). I find myself sometimes confused on how to edit structures. For such modeling work, I tend to use SYBYL or MOLDEN.

I feel that UCSF Chimera has lots to offer, and we are using this program in Drug Design course with students.

These are all 3D viewers. In my experience, in many practical industrial application areas, the capability to comfortably deal with a 50K records 2D library SD file is a more frequently encountered problem than reading a single 3D PDB structure and making a pretty picture from it….

I use Jmol because it is very web friendly (you can export views to live web pages), making it a great teaching tool. Being a Java application it is easier to maintain (at least until, JVMs are no longer supported). All the other options require machine specific compilation. The quality of Jmol’s images has also been improving rapidly and the user interface seems to be more straightforward than the others.

Jonathan Gutow

I use Interchem for all my viewing and interface to other
programs. It has a good user interface (no cryptic icons but good informative text).
WHY? Well I have mostly written it myself, I know wat it is capable of, and if I need new functionality I can easily add it.

Peter Bladon

Coot for looking at maps with molecules – pymol for publication-quality pictures. And whatever text editor on the system to see/tweak whatever’s in the pdb file.

I use YASARA (www.yasara.org) for everything, from visualization to homology modeling and docking. It’s just the coolest (check out the help movies) and fastest around. Even the surfaces adapt fluently when atoms move, and half a million atoms with shadows are also no problem..

While it may not be my favorite (that title is reserved for my own little iPhone application, Molecules), I am stunned by some of the visuals that you can generate using the open source QuteMol client: http://qutemol.sourceforge.net/ . The folks at http://www.nanorex.com have a gallery of molecular machinery renderings generated from this application. They are beautiful.

I teach high school students. And, I like Rasmol because it is the easiest to download and use on our computers. We don’t need IT people to configure any web browsers or allow any permissions to make the system work.

I use MDL Chime (for glancing), PyMOL, and VMD. I’ve enjoyed Jmol when I’ve used it, but haven’t gotten around to making it part of my workflow. It’s interesting to see PyMOL’s creator plug Jmol.

Dr. Delano, I guess you’re mainly plugging open source (and I totally agree with these ideas about the merits), but you also seem to have a genuine liking for Jmol and a desire to promote it. Do I read you right?

This was pretty good in it’s time, as a streaming web3d ( shockwave ) product ( Windows and Mac ).

http://www.candointeractive.com/demos/molecule.htm

It needs a UI update, but the core tech is fast and very accessible.

My vote is for VMD eventhough i never used to PYMOL..i feel vmd is userfriendly. It is my favourite to view the protein-ligand conformation and can save the files in different formats like .wrl.

VMD is easily available and has powerful functions that are useful for molecular modelers.

I can only choose ALL OF THE ABOVE! Every last viewer is valuable, as each has unique
capabilities the others don’t have. Indeed I have all of those you list except VMD installed on my computer for this reason.

I cannot decide for one unique favorite. Depending on the tasks at hand, I use different viewers.
I like jmol for a quick view at structures or scanning over a large number of structures and the jmol as a web applet for presentations, for communicating the key features of a molecule, and for documentation.
I like the ability to just drag and drop a pdb file on the viewer window, or scan across a large number of structures with a simple script, and to combine my notes on the progress of a project with the interactive viewing features of a jmol webpage.
Pymol is great for producing illustrations and high quality animations.
I have so far been using a commercial product for serious modeling and detailed analysis of structures, and I am currently evaluating alternatives.

Chimera for variability of options (the authors have not forgotten that nucleic acids are biomolecules) and making nice pictures.

RasMol for being so super-simple to install and use yet rich in functions.

I use different tools for different jobs. But from the perspective of developing educational resources, I find Jmol to be the best. It is scriptable, extensible, and forkable (though the codebase is, IMHO, complicated). most of all, you can’t beat Jmol’s delivery system (the Web). Jmol as a browser applet means I can offload all of the 3D rendering tasks to Jmol, and use the browser surround for layout, hyperlinks, colors, audio, associated video, etc.

Jmol does have some fairly advanced animation commands, too. you can’t easily produce a traditional movie – you can using javascript! – but traditional movies are linear, and much less useful than short interactive animations and guided exploration anyway. plus Jmol has surfaces, cavities, and other “meta-molecular” features.

the fact that Jmol is free, open source, cross-platform, app and applet, and has active dev and user communities – also attractive.

I use several viewers, but RasMol remains very handy, and once, but only once, all commands and options are mastered, it can give superb figures.
Furthermore, command scripts can be saved for the purpose of reproducing quickly several molecular views.
Very useful for conference and teaching purposes.
Last, but not least, it is rather bugless.

Jmol is good for import/export/web and small piece work, but I like VMD for the big picture.

> but you also seem to have a genuine liking for Jmol and a desire to promote it. Do I read you right?

Absolutely. For technical reasons, Jmol succeeds where PyMOL cannot, and vice versa. Since both packages are under unrestricted open-source licenses, there is nothing standing in the way of interoperability, integration, and codistribution.

As Egon noted, the two packages really should add support for a common molecular graphics data model. That could enable PyMOL-created content to be published robustly via the web, and cound enable Jmol-created content to take advantage of GPUs and ray tracing.

CCP4mg (post 2.0). Very easy to use.

> As Egon noted, the two packages really should add support for a common molecular graphics data model.

That’d be great. Perhaps a seed that could crystallize more widespread interoperability? I see that Egon mentioned developing an open standard (of course). A very exciting idea.

I mean, I don’t envy you the boring standards body meetings ten years hence, but the outset should be pretty fun. And the benefit to scientists the world over is easy to underestimate. I already encourage people to use PyMOL and Jmol; this idea, even while only in a seminal stage, is yet more reason to.

To be fair, both John Stone (VMD) and the Chimera team have also expressed similar desires for an open standard over the years. We have all solved related problems different ways.

But nowadays a common theme has emerged, and whether you call it PyMOL Scenes, Chime Protein Explorer, Jmol Scripts/FirstGlance, ICM Slides, VIDA Bookmarks, or Chimera Demos, we have all more or less converged on common molecular storytelling capability.

So a logical next step would be the development of an open standard capable of capturing enough common value inherent in each of the above solutions to support some level of baseline interoperability: a PDF & PowerPoint-like data model for molecular visualization.

Creation of such a system wouldn’t necessary require participation from everyone, but it would take at least two independent efforts working together, with a judicious 3rd-party champion driving the process. I’m not a big believer in committees – I’d rather see some highly-informed, synthetic, neutral visionary(ies) take on the interoperability challenge as their own personal mission. Start small & conservative, then build on each incremental success…

Given their complementarity and unrestricted open-source status, PyMOL and Jmol are the two obvious starting points. But as this poll & discussion confirms, there are many other packages which matter as well.

I mainly use Chimera to integrate sequence and structure alignments and to make figures for publication. Chimera has many other great tools and features, and the development team is fast, responsive, and always working on adding new features.

I believe that Tom Oldfield (EBI) is developing/has developed an XML view/scene description. It is designed to be portable between programs and is/will be used in AstexViewer.

Warren, going to the ACS meeting? Can you try to meet up with Bob Hanson and discuss an open standard for molecular views?

Honestly, the longer I see this voting mentioned, the more I dislike the freedom of choice.

Where is the vote for the Astex viewer, Avogadro, ICM Browser, COOT, and all the commercial tools from OpenEye, CCG, Schrodinger, Tripos, Accelrys, name it ? Just look at the list of OpenBabel related project and find much more
http://openbabel.org/wiki/Related_Projects

Please add the tools and comparison to the Wikipedia article
http://en.wikipedia.org/wiki/Software_for_molecular_mechanics_modeling

If people do not like mixing ‘molecular mechanics modeling’ with ‘molecular viewing’ then just create another page, e.g.
http://en.wikipedia.org/wiki/Software_of_molecular_viewers

“Honestly, the longer I see this voting mentioned, the more I dislike the freedom of choice.

Where is the vote for the Astex viewer, Avogadro, ICM Browser, COOT, and all the commercial tools from OpenEye, CCG, Schrodinger, Tripos, Accelrys, name it ?”

Joerg- post a list of the viewers you want to see in the poll and I’ll add them.

I agree, the ‘Other’ category is quite large at 10% with 564 total votes.

@Xavier – If the purpose is to collect information, then check the links of my previous comment. There people could also enrich the information already around, if they really want. This is what I would prefer.

If the purpose is to keep ranting about a few viewers (excludes for me modeling and editing tools), well, then at least all freeware viewers should be mentioned, which can be taken from the links mentioned as well, and adding especially the Astex viwer, and COOT.

I like BALLView myself. It has Python scripting, sophisticated rendering, and the learning curve seems shallower than PyMOL’s.

Jmol is very simple to use and very fast to learn.
It is written in java so is highly portable.
Both at school where linux used and at home where pupils still use windows.
Pupils’s favourite feature: 3d molecules in anamorphic.

Giuseppe Stavole
high school chemistry teacher
Bolzano Italy

I use GaussView, gOpenMol, VEGA ZZ.

When I need only visualization, I prefer most of the time to use Raswin because of the following: (a) it is very easy to learn how to use it; (b) it is very intuitive; (c) its graphics are of enough good publishable quality;(d) from the moment you learned its simple but powerful command line language, it is very easy to right scripts for it – though I would like it to have more powerful programming capabilities.
I should say that there are more powerful viewers, but the compactness and ease of use of Raswin are its strength and beauty.

Moshe Goldstein, PhD candidate
The Hebrew University of Jerusalem

The percentages reported by the poll are a little strange. They round to the nearest integer (even through they report the tenths decimal place) and they round down. So… 0.9% would be reported as 0.0%. The order is correct, meaning that MOLDEN ranking above DSVisualizer means that either MOLDEN has more votes or is tied with DSVisualizer.

[...] the success of our previous poll which was followed by an extensive and interesting discussion, and in order to better our [...]

I voted for Pymol but I use MOLMOL just as frequently. Visualizing a bundle of NMR structures, MOLMOL is easier. What’s happened to MOLMOL? Nobody uses it anymore?

The choice of a viewer or modeler is a difficult question. The answer depends on what sort of a user we have. As I saw one day, a chemist studying protein-ligand docking / steric restrictions was delighted by the not too well known PV protein modeler. On the other hand, PyMOL could be the best choice for presentations.

I’ve always used VMD, but I have to admit I’m a little dissapointed with its ability to superimpose structures that don’t have a 1-to-1 residue match.

There is one package, called Quanta, which doesn’t appear on the list and doesn’t seem to be very popular. Its a great little tool for manipulating structures – changing rotomers and backbone conformations is easy and manual. It is easy to insert or remove residues, with algorithms to clean up the structure afterwards, plus it interfaced with CHARMM nicely.

I usually use Pymol or VMD. VMD has great rendering capabilities, while pymol is a bit more user friendly.

Note that OpenAstex is a very nice alternative, especially for web pages (as an applet). It has a powerfull scripting language, similar to that of Jmol but much more easy to use in my opinion.

I used PyMol, VMD, RasMol and YASARA, and stayed with YASARA for years now.
YASARA is extremely fast whilst displaying the coolest 3D graphics around. The different scenes (viewing styles) are adaptable with 1 button and give me an instant clarity on sidechains, fold, etc …
I also added my own menus and commands to YASARA with customizable options by just writing simple (or more complex) plugins in Python.
Love it!

how did the poll come out?

I used PyMol and ICM for research and used to teach using PyMol. I will change for another program since Pymol is not free and opensource anymore since schrodinger took over the licensing.

Thus I will switch back to VMD.

[...] while you’re at it, you might want to revisit our poll on your favourite macromolecular viewer. Share [...]

I use MOE – excellent, and beautiful (includes POV ray rendering).

On the free side I of course love VMD, and QuteMol