Structure Refinement Using XPLOR-NIH

Introduction

Here we describe the protocol for Xplor-NIH refinement of NMR structures.  The Xplor-NIH molecular structure determination package is available at http://nmr.cit.nih.gov/xplor-nih/

The restrained simulated annealing protocol uses many of the updated features of the Xplor-NIH software, including:

• the IVM module for torsion angle and rigid body dynamics
• a radius of gyration term to represent the weak packing potential
• database potentials of mean force to refine against:

1. Cα/Cβ chemical shifts
2. multidimensional torsion angles,
3. backbone hydrogen bonding database 

• RDCs can be included but aren't described here

The topology and parameter files protein.top and protein.par, which were designed to agree with bond lengths and angles from the CSDX force field, are used.

Protocol

Setup

The files needed to run the Xplor-NIH refinement are:

• prot_noe.tbl
• prot_dihe.tbl
• prot_hbond.tbl (optional)
• prot_shiftsCACB.tbl
• prot.psf (described below), can be obtained from mkpsf_prot.inp
• sa_refine.inp (edited to reflect your protein parameters)
• xplor format PDB files to be refined (described below)

The NMR restraints should be in Xplor/CNS format.

If you have restraint files in another format, one way to convert is using PdbStat

The input PDB files should be in Xplor-NIH format

One way to convert them from another format, such as Cyana or CNS is using PdbStat

Creating protein structure file, (psf)

Editing and running the mkpsf_prot.inp script

The X-Plor psf file specifies the atom parameters and connectivity for all amino acids.  The xplor script mkpsf_prot.inp is edited with the protein sequence in three letter code and used to generate the psf files by running:

xplor < mkpsf_prot.inp > temp.out

Download the script mkpsf_prot.inp.
Edit the script with your sequence and to match your input xplor PDB files:

1. Have chain A or no chain ID to match your input PDB file.
2. Add the TOPPAR:toph19.pep line if your input structures have NH³⁺ and COO- at the N- and C-termini with the atoms HT1, HT2, HT3 and OT1, OT2, respectively.
3. Fix the Histidine patch to match the protonation state / tautomer.
4. If you protein sequence starts at residues 43, for example, add extra residues to the sequence and uncomment the "dele sele (resid 1:42)" line by removing the "!".
5. Add cis Proline patch if you have any cis Prolines.
   

Edit and run the simulated annealing script

Edit simulated annealing script

Download the script prot_sa_refine_nogyr.inp (for refinement without the radius of gyration, R_gyr, such as proteins that aren't globular, or you haven't yet determined the residues that should be included in the "core")

or prot_sa_refine.inp to include the R_gyr term.

Edit the script:

1. To match your input PDB and .tbl file names.  Remove the "!" from prot_hbond.tbl if you have hbond restraints.
2. Edit the radius of gyration line for your "core" residues.

For example, if your core residues are 13-92 , then the number of ordered residues, N_^residues is 80.  Use the equation below to determine the radiue of gyration.  In this case R_gyr = 11.6.

    Rgyr = 2.2 * (Nresidues^ 0.38) 

The line in the prot_sa_refine.inp script should look like this:

    assign ( resid 13:92 ) 50.0 11.6

The number 50 is a scaling constant, so don't change it.

Run the Xplor-NIH refinement

Edit the number of structures that your are  refining and run like this:

xplor < prot_sa_refine.inp > sa.out &

Check the sa.out for the line "ERR" to find errors.  Typically there will be errors if the .psf file atom names do not match the input pdb file names.  This can help you track them down.