February 12, 2009

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

New Action Items

• Eric will send Uniprot-PDB mapping e-mail message from PDB-L to Conrad again
• Greg will look into implementing interpolating ribbons and displaying ribbons for structures with only backbone atoms
• Greg will remove lens code from Chimera core
• Conrad will look into semantics and occurrences of SPLIT records in PDB

Previous Action Items

• Conrad will look at how to get annotation information from RCSB that is not in the PDB or MMCIF files.
• Scooter will look into providing non-staff access to trac -- will be done Friday
• Everyone will look at open trac issues and assign milestones if possible -- still working through them
• Scooter and Eric will create and account for Wei to access trac -- depends on completing non-staff access to trac
• Conrad will look into disabling stereo in orthographic view
• tef will contact Adobe for more information about embedding 3d graphics in PDF documents -- add as a low priority enhancement request to trac

Minutes

1. Previous Action Items
2. Open bug report updates
   • Elaine mentioned that displaying ribbons for CA- or P-only structures is still being requested by many. Currently, this is unimplemented because there is nono ribbon orientation information when only backbone atoms are present. Greg will look into implementing a non-oriented ("worm"-style) ribbon using only backbone atom coordinates. Conrad suggested that, eventually, the orientation information can be computed from helical and strand axis orientation, eg ribbon is oriented parallel to helical axes and perpendicular to strand axes.
   • Elaine also mentioned demand for ribbons that pass through (or at least nearer) the actual atom positions. Conrad claimed that the ribbons looked ugly when he investigated this a while back. TomG mentioned that he has code (from computing flight trajectories) for selecting tangents that minimize the "wiggle" in the ribbon. Greg will look into implementing an interpolating spline, eg Cardinal spline, while incorporating TomG's changes.
3. Discussion of sequence annotation possibilities -- Eric
   • Eric mentioned that he will be installing ZSI (a web service Python package) so that he can use it for accessing Uniprot for additional sequence annotations.
   • Conrad requested that Eric resend the PDB-L message describing Uniprot-PDB mapping.
4. Removal of unused features
   • Conrad requested that this be postponed since he did not have a chance to look at minutes from previous meeting.
   • TomF suggested that the lens code be removed from Chimera core to simplify maintenance (and possibly get some performance improvement). Greg and Conrad estimated that the task requires 1-2 days if no renaming of "lens" in the code is required. Greg will do the following:
     • Remove blending lenses but keep opaque lenses because the background lens is an opaque lens
     • Remove attribute lookups from opaque lenses
     • Keep only the "background" instance of opaque lenses
     • Measure performance impact
5. PDB SPLIT record
   • Eric mentioned that there are other new records in PDB v3.2.
   • Greg suggested that Chimera be modified to recognize new records so no "Unknown record" messages are emitted for v3.2 files
   • Conrad wanted to have SPLIT records interpreted so that complete models are read rather than displaying only partial models
   • TomG suggested that we check the semantics and frequency of occurrence of SPLIT records and then decide what to do
   • Conrad will try to answer the following questions:
     • How often do SPLIT records appear in the PDB?
     • What does the mmCIF equivalent to partial models look like?
     • Can chain ids be reused in different partial models?
6. Meeting ends. Subsequent topics not discussed.
7. Initial model appearance