Re: [phenixbb] centered triclinic lattice

Thanks Pavel. I tried adding _struct_ncs_oper matrices to the input cif file but got a phenix error that an identity rotation with non-zero translation elements was improper. It would be great if the code could be modified to accept this transformation.

On Dec 14, 2022, at 11:25, Pavel Afonine wrote:

Hi Cathy,

making the 2nd copy and working in P1 seems as fine option to me. And no, cctbx doesn't know A1.

If this is for refinement you can use NCS constrains which will enforce the strict symmetry on everything: coordinates, occupancies, B factors.

Pavel

On 12/14/22 07:31, Cathy Lawson wrote:

...

Looking for suggestions — cctbx/phenix seems not to be able to handle a triclinic lattice with centering.

Specific Example: PDB entry 1LKS with cell parameters/space group “A 1”: CRYST1 31.970 65.080 27.250 94.57 111.56 82.98 A 1 2

A 1 diagram: http://img.chem.ucl.ac.uk/sgp/large/001bz1.htm

Are there any other options for working with this data in phenix other than creating an explicit 2nd copy of the coordinates at x, 1/2+y,1/2+z and setting space group to primitive lattice P 1?

Thanks in advance.