Hi Dan,

There is currently no Chimera capability for moving atoms other than rotating about bonds. But I have been wanting to add mouse modes for moving and rotating sets of selected atoms around. Here is some Chimera code

http://www.cgl.ucsf.edu/home/goddard/temp/moveselection.tar

that adds a mouse mode to move selected atoms in the screen x and y directions with the mouse. The archive contains a directory called MoveSelection. Put it in your Chimera distribution in

chimera/share

or on Mac in

Chimera.app/Contents/Resources/share

Then when you start Chimera use menu entry Favorites / Preferences, and choose category "Mouse" from the preferences dialog. This shows the mouse modes table with the available modes along the top line. The last one is for moving selected atoms. Reassign a mouse button (say button 2) to this mode and you can move selected atoms.

There is no undo.

Tom

Hi Dan,

The only model building capabilities current in Chimera are commands addaa and swapaa which add and swap amino acids, and bond rotation capabilities. We think creating capabilities for building models into density maps like the program O is too big a project. Building models without reference to maps may be easier but I suspect molecular dynamics capabilities or at least energy minimization are needed to make that useful. Chimera does not do MD calculations or energy minimization and again that is a big project we have been reluctant to take on so far.

If you think specific new model building capabilities would be helpful given the above limitations, explain them and we will be happy to consider them.

Tom

Hi Tom,

If you think specific new model building capabilities would be helpful given the above limitations, explain them and we will be happy to consider them.

For the beginning, I think the following "model building" capabilities would make Chimera even more useful (some of them are maybe already implemented, I do not know Chimera so well yet):

1) Move atom, move selection of atoms, delete selection/atom. Add atoms. 2) Change atom to another one. 3) Add/attach functional groups from library 4) I work in the field of nucleic acids, it would be nice one could build the "ideal" structure (B-DNA, A-DNA, Z-DNA) from sequence (if needed, I can provide a references to papers publishing geometry parameters). Possibly change of pseudorotation angle in deoxyribose and ribose.

I am sure other Chimera users would take a benefit if them too. Of course you're right more sophisticated capabilities would probably require force field implementation, and I understand this is (unfortunately) not the direction Chimera is heading to.

Regards Dan

Yes, we have gotten a lot of requests for structure-building features, so we will put some effort into that in the near term. Certainly the simple geometry-based features would come first, but even the more sophisticated force-field type features are becoming more feasible. We already include MMTK under the hood in Chimera for its NetCDF trajectory-reading capabilities, and with some effort we could leverage its minimization features (most effort being getting charges onto atoms and getting Chimera's molecule data into and out of MMTK format).

One simple thing requested at the workshop was the ability to change a bond's length. That's really easy and I'll be doing that first, so if anyone else would want that then let me know and I'll send you the code when it's ready.

--Eric

On Nov 18, 2005, at 8:22 PM, Thomas Goddard wrote:

Hi Dan,

Those are good specific suggestions. Eric Pettersen is the most likely Chimera developer to work on this. I focus on Chimera tools for large molecular assemblies (viruses) and density maps. Eric's to-do list is very long but many people have expressed interest in model building tools including participants at a 2-day Chimera workshop that ended today. So Eric will be evaluating the priority relative to other Chimera feature enhancements, and your input helps.

Thanks,

Tom _______________________________________________ Chimera-users mailing list Chimera-users@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-users

Daniel Svozil wrote:

Hi Tom,

...

If you think specific new model building capabilities would be helpful given the above limitations, explain them and we will be happy to consider them.

For the beginning, I think the following "model building" capabilities would make Chimera even more useful (some of them are maybe already implemented, I do not know Chimera so well yet):

1. Move atom, move selection of atoms, delete selection/atom. Add atoms.
2. Change atom to another one.
3. Add/attach functional groups from library
4. I work in the field of nucleic acids, it would be nice one could

build the "ideal" structure (B-DNA, A-DNA, Z-DNA) from sequence (if needed, I can provide a references to papers publishing geometry parameters). Possibly change of pseudorotation angle in deoxyribose and ribose.

For nucleotide polymers, I've wrapped most of X3DNA in Python in my Ensemble.Legacy library. It's highly alpha code, but I've been able to use it to generate X3DNA parameters from PDB files, and adding support for building nucleotide structures with arbitrary X3DNA parameters wouldn't be hard (you give it twist, tilt, roll, etc, and it gives you a PDB structure consistent with those parameters).

If I had an infinite number of monkeys, I could probably wrap all that into a Chimera extension that let you build arbitrary structures for DNA/RNA, etc with a nice interface. However, I don't, so all I can do is make the code available on a subversion repository and provide minimal advice on how to build onto it.

Dave