[phenixbb] C2 twinning refinement

[whittle]

Hi all-

I'm hoping for some guidance on how to determine the correct twin law,
or possibly re-index my data to allow proper refinement.

The crystal appears to be R32, R3, or C2. Maps seem best in C2, even
though it merges equally well in all these spacegroup options, but R
stays high (~0.36/0.4) and a un-modelled helix visible in an untwinned
3.3 A dataset has no difference density in this set.

My attempt to treat the twinning have not worked (see output, including
cell dimensions, below).

Thanks in advance for any help!

James


phenix.xtriage flags it as most likely twinned with the following
output:

...

Unit cell: (185.85, 120.972, 111.05, 90, 119.157, 90)
Space group: C 1 2 1 (No. 5)
Systematic absences: 0
Centric reflections: 1784
Resolution range: 28.8717 2.51224
Completeness in resolution range: 0.952773

...

Using data between 10.00 to 3.24 Angstrom.

Determining possible twin laws.

The following twin laws have been found:

----------------------------------------------------------------------------------------------------------------
| Type | Axis   | R metric (%) | delta (le Page) | delta (Lebedev) |
Twin law                                  |
----------------------------------------------------------------------------------------------------------------
|  PM  | 2-fold | 0.345        | 0.219           | 0.003           |
-1/2*h+1/2*k+l,1/2*h-1/2*k+l,1/2*h+1/2*k  |
|  PM  | 2-fold | 0.345        | 0.219           | 0.003           |
-1/2*h-1/2*k+l,-1/2*h-1/2*k-l,1/2*h-1/2*k |
----------------------------------------------------------------------------------------------------------------
M:  Merohedral twin law
PM: Pseudomerohedral twin law

  0 merohedral twin operators found
  2 pseudo-merohedral twin operators found
In total,   2 twin operator were found



The present crystal symmetry does not allow to have the its lattice
symmetry expressed in the setting desired.
Becasue of this, a full coset table cannot be produced. Working with the
data in the reduced cell will solve this.

Splitting data in centrics and acentrics
  Number of centrics  : 961
  Number of acentrics : 30683



phenix.refine with 

  twinning {
    twin_law = "-1/2*h-1/2*k+l,-1/2*h-1/2*k-l,1/2*h-1/2*k"
  }

results in the error:

Traceback (most recent call last):
  File "/usr/local/phenix-1.3-
final/phenix/phenix/command_line/refine.py", line 5, in <module>
    command_line.run(command_name="phenix.refine", args=sys.argv[1:])
  File "/usr/local/phenix-1.3-
final/phenix/phenix/refinement/command_line.py", line 89, in run
    call_back_after_monitor_collect=call_back_after_monitor_collect)
  File "/usr/local/phenix-1.3-final/phenix/phenix/refinement/driver.py",
line 1081, in run
    call_back_after_monitor_collect = call_back_after_monitor_collect)
  File "/usr/local/phenix-1.3-
final/phenix/phenix/refinement/strategies.py", line 190, in
refinement_machine
    twin_law = sgtbx.rt_mx( params.twinning.twin_law )
RuntimeError: cctbx Internal
Error: /net/anaconda/scratch1/phenix/phenix-1.3-
final/cctbx/sgtbx/rt_mx.cpp(230): Unsuitable value for rational rotation
matrix.