Rolf Huehne wrote: >Bob Hanson wrote: > > >>Jmol users, exciting news on three fronts: >> >> >>load FILTER >>selection of biomolecules from PDB files >>thermal (and otherwise) ellipsoids and spheres >> >> >> > >I have tested loading of biomolecules again with the extra-large entry >'1m4x'. With revision 9356 (application, Java 1.4.2_03) the biomolecule >looked fine now (in contrast to revision 9339). > >I noticed two things while testing it: > >1) It is almost impossible to apply any commands using the console if >there are about 2,000,000 atoms. After loading '1m4x' with filter option >"*.CA; BIOMOLECULE 1" I tried to apply command "color chain". It just >finished its work after 4 hours... >In contrast it took only about 1 minute to apply the corresponding >command using the popup menu. > > > Right -- it's a problem with the undo/redo. When you give a command in the application, Jmol automatically saves the state. Unfortunately, with these huge systems, that saving is a very time-consuming operation. I set it so that if the save takes more than 1 second, then Jmol automatically disables UNDO. But that's a hack reaction to the problem, not a solution. You can initially disable UNDO using
set UNDO off # (undocumented) and then you won't have that problem. I'm still figuring out a way of doing this elegantly. >2) There is a huge difference in load time and memory usage between >loading the biological unit file for '1m4x' provided by the PDB and >loading the asymmetric unit file with the biomolecule filter option: > > a) biological unit file, all atoms, 1680 models > b) biological unit file, only CA atoms, 1680 models > c) asymnetric unit file, filter "BIOMOLECULE 1", 1 model > d) asymmetric unit file, filter "*.CA; BIOMOLECULE 1", 1 model > > | No. of Atoms | Memory | Load Time > ----|-------------------------------------- > a) | ~16,000,000 | >120 GB | >3 hours > b) | ~2,000,000 | 83 GB | 30 min. > c) | ~16,000,000 | 19 GB | 55 min. > d) | ~2,000,000 | 2.5 GB | 5 min. > > > Say, Rolf, when you tire of that machine, please send it to me. :) Amazing. >(The listed memory usage is the usage of the whole JavaVM immediately >after loading the structure, before any commands were applied.) > >I am wondering about this huge difference in memory usage (~40 times) >because all atoms can still be manipulated separately in case c) and d). > > > You mean the difference between (a) and (c), or between (b) and (d)? Very interesting. Must be the effect of having 1680 models relative to 1. But this sure looks like a bug to me. I'll see if I can track it down with a simpler system. Bob >Regards, >Rolf > >------------------------------------------------------------------------- >This SF.net email is sponsored by the 2008 JavaOne(SM) Conference >Don't miss this year's exciting event. There's still time to save $100. >Use priority code J8TL2D2. >http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone >_______________________________________________ >Jmol-users mailing list >[email protected] >https://lists.sourceforge.net/lists/listinfo/jmol-users > > -- Robert M. Hanson Professor of Chemistry St. Olaf College Northfield, MN http://www.stolaf.edu/people/hansonr If nature does not answer first what we want, it is better to take what answer we get. -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900 ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Jmol-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/jmol-users
