Hi Yufan,
I suggest you generate CIF for each individual non-standard item in your PDB file first. If you are using recent Phenix it should link them together automatically (link_ligands=True is responsible for this, and it is the default).
Make sure it linked them all correct. It relies on input model geometry, so if geometry is suboptimal it may not link them correctly.
Alternatively, you can use custom bond restraints that allow you to link any pair of atoms, any number of pairs:
You can compose a file like shown in above link and give it to phenix.refine GUI or command line, whatever you use.
LINK records are not recognized and used by phenix.refine.
Let me know if you have any questions or need help!
Pavel
Dear Phenix Developers/Users,
I have a question regarding how to generate a correct restrain for a whole nonstandard peptide and how do run refinement correctly.
The nonstandard peptide in my structure is pepstatin A (IVA Val Val STA Ala STA).
I manually put the pepstatin A in the density and generated the cif files for IVA and STA and run the refinement with them.
After refinement, the whole pepstatin A explored as pieces, which I guess the program treated the pepstatin A as single amino acids.
By the way, I also manual input the “LINK” information in the input pdb file as below,
LINK C VAL L 3 N STA L 4 1555 1555LINK N ALA L 5 C STA L 4 1555 1555LINK C ALA L 5 N STA L 6 1555 1555
And also, when I use the “Real Space Refine Zone” by clicking two atoms for the whole pepstatin A in coot, it exploded again, although I input the same cif file in coot as well.
The data quality is not good (2.9 A with CC1/2=12), thus it is difficult to place amino acid individually to fit the destiny.
In the PDB, there are several structures containing the pepstatin A, such as 4TVG, 4AUC. They have the pepatatin A sequence as VVVXAX, but none of them lists the constrains.
Would you please help me and let me know how can I continue with refinement? Please let me know if you need more information.
Thank you vey much!
Yufan Wu
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]