Multistart simulated annealing refinement
Hi All, I'm sure this has already been addressed but is there a way through the PHENIX GUI to carry out multistart simulated annealing refinement either with different starting seeds (velocities) or temperatures automatically or does one have to start a fresh run each time? Thanks! Joe P.S. Similar procedure as described in the PNAS paper: "Multistart simulated annealing refinement of the crystal structure of the 70S ribosome" ___________________________________________________________ Joseph P. Noel, Ph.D. Investigator, Howard Hughes Medical Institute Professor, The Jack H. Skirball Center for Chemical Biology and Proteomics The Salk Institute for Biological Studies 10010 North Torrey Pines Road La Jolla, CA 92037 USA Phone: (858) 453-4100 extension 1442 Cell: (858) 349-4700 Fax: (858) 597-0855 E-mail: [email protected] Web Site (Salk): http://www.salk.edu/faculty/faculty_details.php?id=37 Web Site (HHMI): http://hhmi.org/research/investigators/noel.html ___________________________________________________________
Hi Joe, to do this you can simply compose a script in whatever language you prefer, which prototype looks like this: for random_seed in random_seeds: phenix.refine model.pdb data.mtz main.simulated_annealing=true main.random_seed=random_seed output.prefix=str(random_seed) where random_seeds is the array of random seeds, containing ~100 elements. If you have multiple CPUs then you can obviously split it among them. I'm not sure we have GUI for this and I guess this kind of task is best run from the command line anyway. Pavel. On 12/8/10 10:47 AM, Joseph Noel wrote:
Hi All,
I'm sure this has already been addressed but is there a way through the PHENIX GUI to carry out multistart simulated annealing refinement either with different starting seeds (velocities) or temperatures automatically or does one have to start a fresh run each time?
Thanks!
Joe
P.S. Similar procedure as described in the PNAS paper: "Multistart simulated annealing refinement of the crystal structure of the 70S ribosome"
On Wed, Dec 8, 2010 at 11:03 AM, Pavel Afonine
I'm not sure we have GUI for this and I guess this kind of task is best run from the command line anyway.
We definitely don't have a GUI for this. It wouldn't be terribly difficult to write one, and it could be automatically parallelized without too much extra effort. The problem is that (as Pavel just explained to me) there are many different possible ways to combine the output maps, and it isn't clear which would work best. So it's potentially non-trivial to test it thoroughly. Just replicating exactly what Korostolev et al. describe might be easier. However, my gut feeling is that the result will be similar to kicked maps - although without the SA-derived estimate of disorder. -Nat
participants (3)
-
Joseph Noel -
Nathaniel Echols -
Pavel Afonine