Opened 10 years ago

Closed 10 years ago

Last modified 10 years ago

#127 closed defect (fixed)

Opening 1jj2 pdb format is slow

Reported by: Tom Goddard Owned by: Eric Pettersen
Priority: major Milestone:
Component: Performance Version:
Keywords: Cc:
Blocked By: Blocking:
Notify when closed: Platform: all
Project: chimera

Description

Opening 1jj2 in pdb format takes 6 seconds, while a 3 week old (July 15) Chimera 2 it opens and displays in 2 seconds. Also opening mmCIF 1jj2 in the current Chimera 2 takes only about 1 second.

Timing statements show that the atoms.radii call in structure.py is taking 3.5 seconds for pdb format 1jj2 and only 0.15 seconds for mmCIF. So part of the slow open seems to be atom radius computation.

Change History (2)

comment:1 by Eric Pettersen, 10 years ago

Resolution: fixed
Status: newclosed

When residue names were switched over to fixed strings, PDBio was no longer canonicalizing them correctly (because a reference wasn't being passed), so two-letter or shorter residue names (nucleic acids) were not matching their templates and many more rings/atom types were being computed.

comment:2 by Eric Pettersen, 10 years ago

Component: Molecular DataPerformance
Note: See TracTickets for help on using tickets.