Hi Lars Currently, Chimera does not have any functionality that interfaces directly with MySQL-type relational databases, although this will eventually be implemented as part of another project (but not for a while still:-). However, Python *does* have several modules devoted to accessing databases, and one could write an extension to Chimera to accomplish this. We would certainly be willing to assist you in the Chimera-side development of such an extension if this is the approach you ultimately decide to take. See here for more information: http://www.python.org/topics/database/modules.html I think that all three approaches you detail are very reasonable. There are generally two ways to create a model in Chimera. The first is by reading in a known file format (such as PDB or mol2), and the second is by manually building the model from raw information. It seems that your (1) and (3) take the first approach - the server side (i.e. your database) is responsible for processing the raw information and giving Chimera a file in a format it understands. Your method (2) takes the other approach, by supplying Chimera with raw information, which it then builds into a model. This is also possible in Chimera, although this all happens in the C++ code, which is not distributed with Chimera. We don't currently have any examples illustrating this in our Programmers Guide, but if you decided to pursue this route, we could provide you with some instructional examples on how to do this. It is hard to say which one of these approaches is best, because it depends on your aims - speed/efficiency? ease of implementation? Dan Greenblatt Computer Graphics Lab University of California, San Francisco dan@cgl.ucsf.edu

first)

My plugin accesses the db and creates a temporary file - then it tells chimera to load it. Advantage chimera can use it native (and probably fast) reading routines...

second)

The plugin fetches the data form the db and builds up the molecule in chimera by use of the python-interface...

third()

I tell chimera to access the database X, use password Y, and run select / or call a stored procedure. the stored-procedure or (for older mysql a linked extension function) could build up an input-format chimera understands from the actual data. the select-version would be a bit ugly cause we would need several of them or place some file-blobs in the bd.

I hope your are interested to do some thing about approach three...

Sincerely Lars _______________________________________________ Chimera-dev mailing list Chimera-dev@cgl.ucsf.edu http://www.cgl.ucsf.edu/mailman/listinfo/chimera-dev