PdbStat

From NESG Wiki
Revision as of 17:00, 30 October 2009 by Admin (talk | contribs)
Jump to navigation Jump to search

PdbStat

The PdbStat program, written and actively developed by our colleague Roberto Tejero (Universita de Valencia), is routinely used in the Montelione lab for the analysis, conversion, and manipulation of coordinate and constraint files for protein structure determination. The program is also an integral component of the Protein Structure Validation Software package (Ref. 1) used across the NESG consortium.

The current version of the software is PdbStat 5.1 (July, 2008). There is no separate publication for PdbStat; we cite this program using Bhattacharya et al, 2007 (Ref. 1). Also, there is no official manual for the program. Below are a number of basic stand-alone PdbStat commands and common uses. Going forward, we can continue to add to this document as more applications are developed.

Anyone interested in obtaining the latest version of PdbStat can contact Roberto at: roberto.tejero@uv.es

COMMANDS AND APPLICATIONS

I. Starting the program

    pdbstat

II. Menu of Commands

menu gives a list of commands and keywords recognized by PdbStat (also get this using the command: help command)

Here are the recognized PdbStat commands for version 5.1: PdbStat> menu

Commands/Keywords currently recognized by PDBSTAT

The commands/keywords currently recognized by PDBSTAT are given below. 
Type "help <command>" for more information on each PDBSTAT function.
  align     analyze     author      bye         class
  clear     close       chain       check       chiral
  delete    debug       dump        energy      evaluate 
  find      fit         fix         get         help
  history   hydrogen    hyper       kabsch      ident/chain
  info      initialize  log/change  load        menu
  missing   order       phi         quit        rama
  read      relax       reset       restore     rotate
  save      see         set         show        rmsd
  to        trans       version     what        write
 homology commands: 
  get      change       homology    homa
 Other kind/commands/keywords of help
  codes(amino codes)    amino(amino geometry)

III. Help documents for specific commands

Each command has a short help document describing its usage and options.
Syntax:
help {topic} {subtopic}

Example: PdbStat> help read

Read Syntax: rea[d] {<arg1>} {<arg2>} {<arg3>}

        arg1 can be:  coo[rdinates] or con[straints] or seq[uence]
        arg2 can be:  pdb or con[gen] or cha[rmm] or dis[man]
        arg3 can be:  the name of the file.
        Examples:   read coord pdb filename.pdb
                    read seq filename
                    read cons congen filename
                    read cons diana filename
Read coordinates/constraints/sequence in specified format from file. This 
command is able to open the file read it locating the number of structures 
stored and then read them.
The user needs to type only <read> then a interactive cycle with questions
begins. The program offers the different formats available, so after read
is typed the program offers <coords/cons/seq> and then after that the
program offers the different formats <pdb,disman,....> 


IV. Preparation of final CNS coordinates for PDB deposition

Before we deposit our final coordinates we would like to order our models in order of lowest conformation energy. Also, we want to perform backbone superposition of the coordinates using ordered [S(phi) + S(psi) > 1.8] residues only. Finally, we want to rotate the ensemble to a desired orientation, which will be the orientation that appears when a user downloads out coordinates from the Protein Data Bank (www.rcsb.org). The final coordinates are saved in the original (CNS) format and IUPAC format for RPF analysis (Ref. 2).

To prepare the final coordinates file for PDB deposition use the following protocol:

PdbStat:

  rea coo pdb [filename]	             #read file with concatenated CNS pdb files
  all					                    #select all the models
  class				                     #classify the models by energy
  order 0.9			                    #determine ordered residues; phi/psi cut-off 0.9
  rmsd best backbone	        	      #backbone rmsd
  [return]			                     #creates an rmsd output file
  write coo pdb [overlayed file]		  #write overlayed coordinates 

Next, open the overlayed coordinates in Molmol and get the desired rotation. Use writetransform command to write the rotation matrix to a file.
Back in PdbStat:
rea coo pdb [overlayed.pdb] #read the overlayed coordinates (all models)

  rotate file [filename]		          #apply rotation matrix
  write coo pdb [final.pdb]		       #write new coordinates
  to iupac				                  #converts atom nomenclature to IUPAC
  write coo pdb [final_iupac.pdb]       #write IUPAC coordinates for RPF analysis

V. Selecting specific models / residues / atoms.

A powerful option in PdbStat is the ability to select specific residues and/or atoms for further analysis. This is extremely useful for superposition and RMSD evaluation of selected residues/atoms.
Syntax and Examples:

   #syntax
      sel[ect]  {model(s)}    {residue(s)}   {atom(s)}                
   #select backbone atoms of residues 5-25,30-50 in models 1,3-5,7-10
      sele  1,3-5,7-10  5-25,30-50  backbone           
   #select N,C,CA,O atoms of residues 5-50 in all models 
      sele  *  5-50  n,c,ca,o                                  
   #combine selection with superposition  
      sele  *  5-50,60-85  *
      rmsd sele backbone 


VI. Conversion of coordinate and constraint formats

In the course of a structure refinement we regularly have to convert between different coordinate and constraint formats for different structure programs (i.e., CYANA, XPLOR/CNS, ECEPP). We routinely use PdbStat for this.
Examples:
# converting CYANA to XPLOR/CNS coordinates:

  rea coor pdb [CYANA.pdb]             #read CYANA models (all)
  to xplor 		                      #converts to xplor; fixes stereospecfic labels 
  write                                #answer questions
       > WRITER: Output file name ?:_  4cns.pdb
       > WRITER: Coords, constraints, aco or sequence file? :_  coor
       > WRITER: ... backbone, heavy, full set? (back/heavy/all/select): all
       > WRITER: What model ?_ : all
       > WRITER:  --  All models to be written 
       > COORD_writer: Format  (pdb/congen/RasMol) ? : pdb	
  1. converting CYANA to XPLOR/CNS distance constraints:
   rea coor pdb [CYANA.pdb]       	  #read CYANA models; you have to read in a pdb or sequence file before the constraints
   rea cons cyana [CYANA.upl]          #read CYANA upls file
   write 				#write to CNS format; add 10% to upper bound;make lower bound
van der Waals (1.8 A)
       > WRITER: Output file name ?:_  4cns_noe.tbl 
       > WRITER: Coords, constraints, aco or sequence file? :_  cons
       > WRITER_constr: Output format 
[congen||discover||ecepp||disman||dyana||cyana||diana||xplor||cns]? : cns
 		> XPLOR_writer: percentage range for upper bound (upp+range)?: 10
 		> XPLOR_writer: range for lower bound (low-range)?: vdw
 		> XPLOR_writer: ... writing ... wait 
  1. converting CYANA to XPLOR/CNS dihedral constraints:
    rea aco cyana [CYANA.aco]
    write
        > WRITER: Output file name ?:_  4cns_dihe.tbl
        > WRITER: Coords, constraints, aco or sequence file? :_  aco              
> WRITER_constr: Output format (congen || discover || impact || disman || diana || xplor || cns)? : cns


VII. Constraint Violations

One can analyze constraint violations using PdbStat. It is best to convert the coordinates and constraints to IUPAC format; this is the approach used internally within PSVS.
Commands: rea coo pdb [filename] to iupac #convert coordinates to IUPAC format rea cons [format] [filename] #read distance constraints; specify format noe to iupac #convert distance constraints to IUPAC rea aco [format] [filename] #read dihedral constraints; specify format set cutu 0.2 #set cut-off for distance violations to 0.2 A see viol noe [sum/ave/center] #see noe violations using sum/average/center averaging see cutaco 1 #set cut-off for dihedral violations to 1 deg. see viol aco #see dihedral violations above threshold

VIII. Sorting of Distance Constraints

There are a number of options in PdbStat for sorting or culling distance constraints.
cons clean #keep conformationally-restricting constraints; also removes duplicates noe analysis #NOE statistics (as in PSVS) noe delete intra #delete all intra NOE constraints {| border="1" |- noe keep long #keep only long range (||i-j|| >/= 5) NOE constraints |} noe keep ilv #keep NOE constraints consistent with ILV labeling

IX. Commands for Obtaining Various Metrics

The “eval”, “show”, and “see” commands allow one to evaluate several types of metrics in a structure or ensemble.
eval [procheck/rama] #get Procheck/Ramachandran statistics for model(s) eval dist * 68 sg 119 nd1 #get 68-SG to 119-ND1 distance across all models {| border="1" |- eval [phi || psi || omega] 1 #get phi/psi/omega torsion angles in model 1 |} see lib [residue] #see library definitions for residue type select 1 119 * #use select and see in tandem see coo #coordinates for residue 119 in model 1

X. Other Functions

# renumbering / resetting residue numbering reset * 10 #sets first residue in file to 10 reset * #resets coordinates to original # ordering ensemble using FindCore algorithm (Ref. 3): {| border="1" |- find [ -bb || -heavy || -all || -noe ] |} # contact map generation based on coordinates or constraints contact > DRAW_cntct: from coordinates or constraints (coor/cons) ?:_ coor > MAIN_cntct: What model do you want (1-20) or average ?_ :ave > do_average_coords(): Making average for backbone atoms > do_average_coords(): Calculating center of masses > do_average_coords(): Calling optimal rotation for backbone > do_average_coords(): Calc. average coordinates backbone > MAIN_cntct: distance cutoff ?:_ 4.5 > MAIN_cntct: atom type (hydr, heavy, all) ?:_ hydr #writes a postscript file


REFERENCES

1. Bhattacharya, A., Tejero, R., and Montelione, G. T. (2007) Evaluating protein structures determined by structural genomics consortia. Proteins 66, 778-795.
2. Huang, Y. J., Powers, R., and Montelione, G.T. (2005) Protein NMR Recall, Precision, and F-measure scores (RPF scores): structure quality assessment measures based on information retrieval statistics. J. Am. Chem. Soc. 127, 1665-1674.
3. Snyder, D. A. and Montelione, G. T. (2005) Clustering algorithms for identifying core atom sets and for assessing the precision of protein structure ensembles. Proteins 59, 673-686.

-- Main.JimAramini - 14 Aug 2008