Hi Greg,
I discovered that at least one of the old mmCIF example files 1voq_1vor_…1vp0.cif from the RCSB has numbers such as coordinates right justified instead of left justified as in the mmCIF files you download from the normal PDB entry pages.
With right-justified numbers and left-justified character fields (e.g. atom name) code to guess the fixed field positions assuming every field is left-justified fails. I don’t see any way to remedy this. But I also suspect these files are obsolete.
Another RCSB location with 9 large mmCIF files is
Cathy Lawson suggested in the email below about 2 months ago that all split PDB entries were going to get single-file mmCIF versions within a month. I believe there are hundreds of split entries. They were supposed to appear in the above ftp directory. Apparently they only put a few there.
Tom
On Mar 6, 2014, at 10:23 AM, Cathy Lawson <cathy.lawson@rutgers.edu
<mailto:cathy.lawson@rutgers.edu>> wrote:
Dear Tom and Tom,
I was asked to write to both of you as a follow up to a discussion at
our RCSB PDB group meeting yesterday.
This is regarding combining of large entry split-entries into single
entry mmCIF files, and the ability of UCSF Chimera to handle these files.
As you are likely aware, wwPDB publicly announced last year that
split entries would be combined, and this work is now in progress.
http://www.wwpdb.org/news/news_2013.html#22-May-2013
When the combined files are ready for release (we think this will be
in the next month or so),
they will be added to "large_structure" subdirectory of the wwPDB
ftp (ftp://ftp.wwpdb.org/pub/pdb/data/large_structures/)
for a grace-period of about six months, after which all split entries
will be obsoleted and large entry cifs will
moved into the main archive.
We are concerned about the impact that these archive changes may have
on the ability of UCSF Chimera to handle
large coordinate file entries. In my experience as well as others in
our group, the 3j3q mmCIF file (complete hiv capsid) in particular is
problematic
to upload/view using Chimera, either crashing or taking many hours to
load. Our current workaround so we can make
pictures is to load a PDB file containing only C-alpha atoms, but
even this is an extremely slow load (say 5-10 minutes, depending on
the machine).
We wonder whether it might be possible to develop a means to more
quickly load and view large structures in Chimera,
e.g., perhaps skipping ribbon calculations and going directly to a
multiscale model surface view?
A shortcut like this would be most welcome by to all of us who
regularly need to view large structures,
and would be similar in spirit to the way Chimera initially displays
large maps with binning.
Can you let me know whether there are plans along this line and if so
when they might be implemented in future releases?
Also if there is any way that we can help? (e.g., several of us would
be keenly interested beta-testers!)
Cheers and Best wishes,
Cathy
******************
Catherine Lawson, Ph.D.
EM project manager for RCSB-PDB
http://emdatabank.org <http://emdatabank.org/>
cathy.lawson@rutgers.edu <mailto:cathy.lawson@rutgers.edu>
USA +1 (848) 445 5494
*****************
Hello, I'm working on improving Chimera's mmCIF support. So if anyone has a mmCIF file that is not from the PDB and that they are willing to share, please send it to me.
Many thanks,
Greg
_______________________________________________
Chimera-users mailing list
Chimera-users@cgl.ucsf.edu
http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users