September 10, 2009

Attendees: Greg, Conrad, Eric, Elaine, TomG, TomF

Action Items

• Eric will change MolGraph so that it uses private containers for atoms and bonds instead of replacing current containers
• Eric will change DBPuppet so that models opened from chimerax files do not have preset automatically applied

Items

1. TomG suggested that molecule.atoms should return a consistently ordered list of atoms throughout the lifetime of the molecule (as opposed to having the atom ordered altered upon events such as ring finding or bond creation). Eric said that the reordering is done when MolGraph recomputes graph components and that molecule.atoms and molecule.bonds need not be reordered by keeping a second set of references to atoms and bonds. This makes molecule.atoms always consistent and molecule.sorted_atoms may then be removed.
2. Remote chimerax scripts are broken in the pre-1.4 snapshot because the initial view is now "smart" instead of white wireframe. Since chimerax scripts customize the molecule display anyway, it was agreed that its better to skip the preset if a molecule is opened during chimerax processing.
3. Conrad reported that the surface code can process the first thousand or so PDB files without throwing exceptions. Elaine commented that any surface replacement code must support area calculations.
4. Conrad mentioned that Patsy wants Chimera images/sessions for SFLD entries. If there will be many entries to process, it will be important to have some (semi-)automated tools. In particular, if a PDB code and a set of active site residue specifiers are available, using a combination of preset, select, focus and align may yield reasonable (starting) views.
5. TomF mentioned that image generation may be needed for GASPS as well. For more information, we need to contact Ben Polacco when he starts with Patsy.
6. Meeting adjourned