In a course on structural bioinformatics I teach usually do a "hands-on" class where the students do structure building (using a high resolution map and Coot) so that they have a feeling about how the PDBs they download are obtained. This year I'd like to make them refine the structure to see how the changes they do are reflected on the statistics (Rfactor, Rfree, etc.). I wouldn't want to install huge packages such as CCP4 or Phenix on all teaching lab computers, so I was wondering if there still exists a standalone software I could use to make them do a simple refinement with their PDBs to see what happens. Any hint?
Hi, nowadays such "things" tend to come in packages. You could go for cns if you think CCP4 or Phenix aren't suitable. For CCP4 though, you do not need all the binaries for simple refinement (and I suppose visualizing the results). You'll need to keep very little in fact, especially if you do not want to recompile any program.
I think you can install refmac or the similar refinement program as a stand alone program use the refmac source code or binary but you have to use the script not GUI
Since refinement entails so may different parameters (restraints & constraints), I think you would do more justice to your students by actually doing refinement with a real refinement program like Refmac, Phenix, CNS on some protein pdb yourself and then take a PrintScreen shot, of let's say Refmac NCS restraints as tight, medium, or loose, and then provide them with a copy of the Printscreen shot and have them analyze the differences on the refined pdb. Also by taking PrintScreen shots of the different other parameters, especially the most common ones (ex. Phenix simulated annealing) that are often modified from the default settings and giving copies to the students would be very beneficial.
- Badges
- Science method