Hello,

I am refining a low resolution data. According to sigma vs. resolution, my resolution is 4.7A.

not sure what is "sigma vs. resolution" but hope you made the right decision.

Because there are four identical molecular per asymetric unit, I can use positional refinement. There are about 65% alpha-helices in the structure, so I want to use secondary structure restrain. Does phenix has this restrain in the refinement?

Sorry for missing documentation on this subject.

First of all, get the latest version of PHENIX (the one you have is very old):

http://www.phenix-online.org/download/nightly_builds.cgi

Then (once you have the latest), yes, you can use secondary structure restraints.

phenix.refine model.pdb data.hkl main.secondary_structure_restraints=true

so phenix.refine automatically determines secondary structure and applies the restrains. Note, GIGO principle applies, that is if your input model is distorted then it's unlikely that the procedure will figure out the secondary structure correctly.

To make it a less black-box, you can run

phenix.secondary_structure_restraints model.pdb > ss_restraints.params

and then check (and edit if necessary) the ss_restraints.params file, and then run phenix.refine as following:

phenix.refine model.pdb data.hkl main.secondary_structure_restraints=true ss_restraints.params

I hope Nat comments more if I missed something.

Let me know if you have any other questions.

Good luck!
Pavel.