Dear all, I'm trying to get DEN refinement going from the command line. The run aborts quickly with the following error in the log file: Sorry: Cartesian simulated annealing should not be turned on for DEN refinement. Annealing settings are controlled in the den parameter scope. The DEN scope in input.def looks like this: den { reference_file = den_ref_Jun.pdb gamma = 0.5 kappa = 0.1 weight = 30 sigma = 0.44 optimize = True opt_gamma_values = 0 0.2 0.4 0.6 0.8 1 opt_weight_values = 3 10 30 100 300 num_cycles = 12 kappa_burn_in_cycles = 2 bulk_solvent_and_scale = True refine_adp = True final_refinement_cycle = False verbose = False annealing_type = *torsion cartesian output_kinemage = False restraint_network { lower_distance_cutoff = 3 upper_distance_cutoff = 15 sequence_separation_low = 0 sequence_separation_limit = 10 exclude_hydrogens = True ndistance_ratio = 1 export_den_pairs = False den_network_file = None } Annealing type is torsion from what I can tell. What am I doing wrong? (Release tag: 1394, Platform: intel-linux-2.6-x86_64 redhat-e6.2) Thanks. Andreas -- Andreas Förster, Research Associate Paul Freemont & Xiaodong Zhang Labs Department of Biochemistry, Imperial College London http://www.msf.bio.ic.ac.uk
Hi Andreas,
My guess is that you have main.simulated_annealing=True also active in your
parameter file. You need to set this to False, and then the refinement
should proceed.
Jeff
On Tue, Jun 11, 2013 at 11:44 AM, Andreas Förster
Dear all,
I'm trying to get DEN refinement going from the command line. The run aborts quickly with the following error in the log file:
Sorry: Cartesian simulated annealing should not be turned on for DEN refinement. Annealing settings are controlled in the den parameter scope.
The DEN scope in input.def looks like this:
den { reference_file = den_ref_Jun.pdb gamma = 0.5 kappa = 0.1 weight = 30 sigma = 0.44 optimize = True opt_gamma_values = 0 0.2 0.4 0.6 0.8 1 opt_weight_values = 3 10 30 100 300 num_cycles = 12 kappa_burn_in_cycles = 2 bulk_solvent_and_scale = True refine_adp = True final_refinement_cycle = False verbose = False annealing_type = *torsion cartesian output_kinemage = False restraint_network { lower_distance_cutoff = 3 upper_distance_cutoff = 15 sequence_separation_low = 0 sequence_separation_limit = 10 exclude_hydrogens = True ndistance_ratio = 1 export_den_pairs = False den_network_file = None }
Annealing type is torsion from what I can tell. What am I doing wrong?
(Release tag: 1394, Platform: intel-linux-2.6-x86_64 redhat-e6.2)
Thanks.
Andreas
-- Andreas Förster, Research Associate Paul Freemont & Xiaodong Zhang Labs Department of Biochemistry, Imperial College London http://www.msf.bio.ic.ac.uk ______________________________**_________________ phenixbb mailing list [email protected] http://phenix-online.org/**mailman/listinfo/phenixbbhttp://phenix-online.org/mailman/listinfo/phenixbb
I do indeed. I thought I control simulated annealing through this? If I turn this off, will I still get simulated annealing, activated through DEN? Thanks. Andreas On 11/06/2013 4:54, Jeff Headd wrote:
Hi Andreas,
My guess is that you have main.simulated_annealing=True also active in your parameter file. You need to set this to False, and then the refinement should proceed.
Jeff
In phenix.refine you control simulated annealing that way, but
phenix.den_refine is a specialized protocol which takes advantage of much
of the phenix.refine machinery while controlling most things through the
'den' scope. It's a bit of a hack, but it was the fastest way to pull a
prototype together. You'll definitely have annealing cycles without turning
them on in the main scope, so long as 'den' is selected as a strategy
(which is is by default).
If you ran the main version of simulated annealing you would add annealing
cycles which didn't take avantage of the DEN network update steps, which is
counter to how each cycle is meant to run.
Jeff
On Tue, Jun 11, 2013 at 12:06 PM, Andreas Förster
I do indeed. I thought I control simulated annealing through this? If I turn this off, will I still get simulated annealing, activated through DEN?
Thanks.
Andreas
On 11/06/2013 4:54, Jeff Headd wrote:
Hi Andreas,
My guess is that you have main.simulated_annealing=True also active in your parameter file. You need to set this to False, and then the refinement should proceed.
Jeff
______________________________**_________________ phenixbb mailing list [email protected] http://phenix-online.org/**mailman/listinfo/phenixbbhttp://phenix-online.org/mailman/listinfo/phenixbb
Ok, that's fine - and it's running now. But I had to fix a number of other things: enable positional refinement, one macrocycle only... When I run phenix.den_refine --dry_run, aren't these inconsistencies and missed requirement supposed to be caught, or do I misunderstand things? Andreas On 11/06/2013 5:37, Jeff Headd wrote:
In phenix.refine you control simulated annealing that way, but phenix.den_refine is a specialized protocol which takes advantage of much of the phenix.refine machinery while controlling most things through the 'den' scope. It's a bit of a hack, but it was the fastest way to pull a prototype together. You'll definitely have annealing cycles without turning them on in the main scope, so long as 'den' is selected as a strategy (which is is by default).
If you ran the main version of simulated annealing you would add annealing cycles which didn't take avantage of the DEN network update steps, which is counter to how each cycle is meant to run.
Jeff
-- Andreas Förster, Research Associate Paul Freemont & Xiaodong Zhang Labs Department of Biochemistry, Imperial College London http://www.msf.bio.ic.ac.uk
Hi Andreas,
The --dry_run flag is designed to evaluate a phenix.refine run,
and admittedly I never tested that with phenix.den_refine. That test is
checking for valide phenix.refine parameters, and your inputs would check
out fine.
In the absence of a parameter file, the den_refine routine makes the
necessary changes to the default scheme (one macrocycle, assures positional
refinement is active, etc.), but a phenix.refine parameter file would
overwrite these and could lead to the catches that you're seeing.
I'll try to make the --dry-run flag work more sensibly with
phenix.den_refine.
Jeff
On Tue, Jun 11, 2013 at 1:21 PM, Andreas Förster
Ok, that's fine - and it's running now. But I had to fix a number of other things: enable positional refinement, one macrocycle only...
When I run phenix.den_refine --dry_run, aren't these inconsistencies and missed requirement supposed to be caught, or do I misunderstand things?
Andreas
On 11/06/2013 5:37, Jeff Headd wrote:
In phenix.refine you control simulated annealing that way, but phenix.den_refine is a specialized protocol which takes advantage of much of the phenix.refine machinery while controlling most things through the 'den' scope. It's a bit of a hack, but it was the fastest way to pull a prototype together. You'll definitely have annealing cycles without turning them on in the main scope, so long as 'den' is selected as a strategy (which is is by default).
If you ran the main version of simulated annealing you would add annealing cycles which didn't take avantage of the DEN network update steps, which is counter to how each cycle is meant to run.
Jeff
-- Andreas Förster, Research Associate Paul Freemont & Xiaodong Zhang Labs Department of Biochemistry, Imperial College London http://www.msf.bio.ic.ac.uk ______________________________**_________________ phenixbb mailing list [email protected] http://phenix-online.org/**mailman/listinfo/phenixbbhttp://phenix-online.org/mailman/listinfo/phenixbb
Dear Jeff, no worries. I assumed the --dry_run would do more than it actually did. This is fine. It didn't take me long to fix the errors that were printed clearly in the log file. Thanks for your help. Andreas On 11/06/2013 6:34, Jeff Headd wrote:
Hi Andreas,
The --dry_run flag is designed to evaluate a phenix.refine run, and admittedly I never tested that with phenix.den_refine. That test is checking for valide phenix.refine parameters, and your inputs would check out fine.
In the absence of a parameter file, the den_refine routine makes the necessary changes to the default scheme (one macrocycle, assures positional refinement is active, etc.), but a phenix.refine parameter file would overwrite these and could lead to the catches that you're seeing.
I'll try to make the --dry-run flag work more sensibly with phenix.den_refine.
Jeff
participants (2)
-
Andreas Förster -
Jeff Headd