On 10/24/16 23:50, Pavel Afonine wrote:

I think "weight=1" may be a troublemaker.. It may be too far from what's optimal (that program determines by default if you don't specify any weight) and that may result in clashes that happen during refinement. Of course, something else may cause this as well.

Pavel

On 10/24/16 20:07, Oleg Sobolev wrote:

...

Hi Julia,

In addition to Dorothee's response to your another e-mail, I would indeed recommend check the model geometry, in particular, bad atomic clashes (where 2 atoms are very close to each other). You can find them in the .geo file which should be created by this point or in log, under "Nonbonded interactions:". The worst clashes will be the first in the list.

Best regards, Oleg Sobolev.

On Mon, Oct 24, 2016 at 3:19 AM, Flis, Julia mailto:[email protected]> wrote:

Hello,

my question is about the Phenix real space refinement. I am using the Phenix version phenix-1.11-2563.

My command looks as follows: phenix.real_space_refine secondary_structure.enable=true Model.pdb VOL.map elbow.Model_pdb.001.cif Model.pdb_ss.eff resolution=3.8 weight=1

My problem is that the refinement suddenly stops during the first refinement cycle:

Model_real_space_refined.log says:

********************** REFINEMENT MACRO_CYCLE 1 OF 5 ************************* ------------------------------------------------------------------------------- Update NCS operators ******************** ------------------------------------------------------------------------------- Update rotamer restraints ************************* ------------------------------------------------------------------------------- Global refinement ***************** END

In the shell I get this :

[...] distance: 184426 - 184921: 3.388 distance: 184428 - 184920: 3.342 distance: 184448 - 184898: 3.373 distance: 184449 - 184897: 3.187 distance: 184491 - 184860: 3.066 distance: 184531 - 184770: 3.203 distance: 184554 - 184746: 3.137

then it stops and the only files generated are Model_real_space_refined.log and Model_initial.geo.

Could you give me an advise? I have no clue what the problem could be.

Thank you,

Julia