refinement of gb1 from a decent starting structure using multiple RDCs

1) start the refinement calculation.

% xplor -smp 2 -py -o refine.out refine.py 

     [ takes about 13 min ]

2) view starting structure ( approximate structure with correct fold )

% vmd-xplor -noxplor model.pdb

  within vmd-xplor:
     to scale
       cntl + move Mouse1 right
     to view as cartoon
       *Molecule --> [model.pdb] (Mouse3) --> Drawing Method --> New Cartoon
	

3) view precalculated final structures

% vmd-xplor -noxplor `getBest precalc/refine_##.sa.stats`

  within vmd-xplor:
    a) display only backbone traces of structures
      *Molecule --> Drawn --> set selection
         type: backbone
    b) fit all structures to the first using Ca coordinates.
      *Molecule --> Drawn --> edit
         the edit form is opened. In the bottom of this window, choose
            fit to 
              [select] --> set to refine_19.sa
                by
                  [all] --> set to CA
                    select
                      *do it

4) generate average structure

% aveStruct -selection "name CA" `getBest precalc/refine_##.sa.stats` > aveStruct.pdb

 *WARNING* coordinates of most side chains will be meaningless.
 This script does not regularize aveStruct.pdb.

  Compare this to the calculated regularized mean structure:

% vmd-xplor -noxplor aveStruct.pdb precalc/refine_ave.pdb

  within vmd-xplor:
    a) display all heavy atoms
      *Molecule --> Drawn --> set selection
         type: not name H* and not resname ANI
    b) color by rmsd from the mean
      *Molecule --> Drawn --> color by --> Beta
       fix the color scale:
      *VMD --> color --> Color Scale --> Method --> BWR
    c) fit all structures to the first using Ca coordinates.
      *Molecule --> Drawn --> edit
         the edit form is opened. In the bottom of this window, choose
            fit to 
              [select] --> set to aveStruct.pdb
                by
                  [all] --> set to CA
                    select
                      *do it



5) analyze pre-existing structures

% xplor -py analyze.py

[note that noe restraints are split into separate classes in the 
 analyze_##.sa.stats file]

6) compute pairwise rmsd 

% ./pairRMSD.py -selection "name C or name N or name CA" \
  `getBest precalc/refine_##.sa.stats`

7) compute RDC alignment tensor given structures

% calcTensor -plot -rdcWeights "1 0.05 0.108" -normType none \
   "tmv107_nh.tbl tmv107_nc.tbl tmv107_hnc.tbl" \
    `precalc/getBest analyze_##.sa.stats`

8) for steric aligning media, the alignment tensor can be calculated
   directly from molecular structure. For this, use calcSARDC:

% calcSARDC -plot -table bicelles_new_nh.tbl \
    `getBest precalc/analyze_##.sa.stats `

9) compute alignment tensor Da, R in absence of structures
    - using a maximum likelihood algorithm

% calcDaRh -seq protG.seq -normtype none -error "1 0.05 0.108" \
    tmv107_nh.tbl tmv107_nc.tbl tmv107_hnc.tbl

10) accessing the built-in help

% xplor -py
                       XPLOR-NIH version 2.35
 
 C.D. Schwieters, J.J.  Kuszewski,       based on X-PLOR 3.851 by A.T. Brunger 
 N. Tjandra, and G.M. Clore 
 J. Magn. Res., 160, 66-74 (2003).       http://nmr.cit.nih.gov/xplor-nih 

 User: schwitrs     on: x86/Linux       at: 28-Jun-07 09:31:03
python> import rdcPotTools
python> help(rdcPotTools)
.
.
.
python> help(open)

HTML documentation:

% pyXplor -pydoc -g

choose
  *pydoc --> open browser

to open a web page with Python interface documentation


11) complete listing of scripts and tables

Python scripts
  analyze.py    - analyze previously calculated structures
  anneal.py     - structure determination from extended random structure
  aveStruct.py  - generate non-regularized average structure
  calcDaRh.py   - calculate Da, Rh in absence of any structure info
  calcTensor.py - calculate alignment tensor using existing structures
  pairRMSD.py   - compute ave. pairwise rmsd between specified structures
  refine.py     - structure determination from good starting guess
  wrefine.py    - script for explicit water refinement of refine structures
  refine_eefx.py- script for implicit water refinement of refine structures

XPLOR scripts
  hardvdwminim_rgyr.inp - regularize average structure so that it has
			  reasonable bonds, angles.
  sa_tmv_bice_rgyr.inp - structure refinement from good starting structure.


Input restraint tables
  bicelles_new_hnc.tbl - RDC tables from bicelle medium- for three experiments
  bicelles_new_nc.tbl
  bicelles_new_nh.tbl
  tmv107_hnc.tbl       - RDC tables from TMV medium
  tmv107_nc.tbl
  tmv107_nh.tbl
  dihed_g_all.tbl - dihedral restraint table (e.g. from Talos)
  hbda.tbl        - backbone hydrogen bond potential table.
  jna_coup.tbl    - J coupling table
  noe_sum_3.tbl   - NOE table

Miscellaneous files
  axis_500.psf - psf files for tensor orientations - used only in XPLOR scripts
  axis_600.psf
  g_new.psf    - pregenerated psf file for gb1
  model.pdb    - starting structure for refinement
  protG.seq    - sequence for gb1
  precalc/refine_*.sa  - precomputed structures
  
