
                              DIRDIF-2007                             
                                                                    
        Crystal Structure Determination by Patterson Methods and      
        Direct Methods  applied to  Difference Structure Factors      
                                                                    
        Paul T. Beurskens,   Gezina Beurskens,   Rene de Gelder,      
        Jan M.M. Smits (Nijmegen),   S. Garcia-Granda (Oviedo) &      
                        R.O. Gould (Edinburgh)                       

------------------------------------------------------------------------ 
File readme.txt   getting started  ===========>  unix          ---------
---------------   Preliminary information for DIRDIF-2007      July 2007
------------------------------------------------------------------------ 

  DIRDIF is the most powerful method to solve crystal structures
  using your chemical knowledge

  Input data: the usual crystal data and observed reflection data, and
  A. ===> for heavy atoms structures: nothing else !
  B. ===> for appr. equal atoms: one or MORE known structural fragments.

Note: in this message we use single file names relating to the contents
      of the file; the actual names used during execution of DIRDIF
      have prefixed names, 
      e.g. with 'dirdif' or with CCODE = name of the structure at hand.
      In printing (dirdif primer, messages, output listing) these
      prefixes often are omitted and may be capitalised;  examples:  
           ORBASE for dirdif.orbase (a global data file)
           CRYSIN for monos.crysin  (a MONOS related data file) .

Getting DIRDIF by ftp
--------------        Prepare to download DIRDIF into your local
                      directory $HOME/dirdif ( ensure that it is empty )
  ENTER: cd $HOME/dirdif 
  ENTER: ftp ftp.sci.kun.nl  ==> for user name ENTER:  anonymous
                             ==> password = your email address
  ENTER: bin
  ENTER: prompt
  ENTER: mget pub/dirdif/DD2007/unix/dirdif*   
  ENTER: quit
  Use "unzip dirdif" to store files into target directory $HOME/dirdif 

Contents of directory dirdif       for unix              
--------

  -  1 text file: this file: readme.txt with implementation instructions
  -  1 directory DDEXE, empty directory for storage of dirdif executable
  -  1 directory FTN  with the fortran program coding (for compilation)
  -  1 directory monos (test case) with 3 files: atlit, crysin and fref 
  -  1 directory DOCS with PRIMER ( plain text files with information )
  -  2 dirdif system files: ORBASE, ORUSER (data for Patterson search) 
  -  1 dirdif calling script file: dirdif

Note: more documents (MANUAL, gallery) can be obtained via FTP with:
      mget pub/dirdif/DD2007/DOCS/*

     
Getting started
---------------

1 Compile dirdif.f (from $HOME/dirdif/FTN/dirdif.zip) and store the
      executable in $HOME/dirdif/DDEXE  (see info in directory FTN).

2 Set path to $HOME/dirdif  and change directory to dirdif/monos

3 Run test case MONOS (monos is compound code name of test structure ) 


  A:  ENTER: cd monos [ use your editor to read the contens of the files
      ENTER: dirdif monos patty [ read some of the contents of file LIS1

  B.        Note: after dirdif has run once, the compound code is known:
      ENTER: dirdif atmod [ to generate a model for patterson search.
      ENTER: s     [ to select a model from ORBASE:
      ENTER: monos [ the code name of the required model for MONOS .
      ENTER: q     [ to quit], and     ENTER: y [ to accept the model.
      ENTER: dirdif orient  [ to solve MONOS again by vector search. 

Aliasses
--------

  It is suggested to prepare the script files dirdifx , dd , and ddx.
  By calling 'dirdifx' , a plot program should be called at the end.
  Instead of 'dirdif' or 'dirdifx' you could use 'dd' or 'ddx' ,
  respectecively, as an abbreviation.

Major calling instructions 
-------------
              The structural data of your compound (name denoted: CCODE)
              are stored in one directory (its name also denoted CCODE).
              Change directory to CCODE and ENTER:
  dirdif CCODE PARM ... (CCODE: use the actual name of the structure)
              Don't forget CCODE at first call for this structure;
              at all following calls the CCODE may be omitted. 
  In this summary of options, alias DD is used and the CCODE is omitted.
  ENTER (from the proper CCODE directory):

dd patty   [ to start a heavy atom structure solution ;
             minimum requirement: about one S or P per 20 C atoms.
dd orient  [ to start Vector Search if a molecular fragment is known:
             the (one or more) fragment(s) is(are) stored in file ATMOD.
dd atmod   [ to generate a simple model, or to check and modify a model
             known from the literature or by molecular modelling.
dd orflex  [ to generate rigid fragments from flexible molecular model
             given in file ATMOD; output: many fragments in ATMOD. 
dd phasex  [ to expand your trial structure, stored in file ATOMS,
             by dual space phase refinement techniques.
dd four    [ to expand your trial structure, stored in file ATOMS,
             by R2-driven Fourier refinement techniques.
dd four norecy  [ to make one more Fourier synthesis with interpretation
                  of your trial structure without recycling the result.
dd bijvoet [ to calculate the 'Bijvoet coefficient' which determines
             the absolute configuration of the structure at hand.

Input files
-----------
        For examles: see input and output files of test structure MONOS.
        A complete description is given in dirdif.primer (1999).
        Remember the present use of unprefixed file names.

crysin, res, ins, xyzn and cif files can be input for crystal data.
fref, shelx, shelxl can be used to input reflection data.
res, ins, xyzn can also be used to input atomic parameters.
atoms contains input/output atomic parameters (fractional).
atmod contains one or more atoms sets for ORIENT .


Output files
------------

crysin is the working file for input crystal data: any change 
       in CELL contents, Space Group, etc. must be given in CRYSIN. 
atoms is output file with new atomic parameters.
lis1 and lis2 are output listings (at present: too much! skip it!).
xyzn is output 'RES' to be used as input to SHELX LS program.
ccode.res is the same 'RES'-file which is input to MERCURY plot program.


Modifications 2007
------------------ 
                   Major changes in DIRDIF-2007 relative to DIRDIF-1999: 
- it is one single program (it was a collection of many programs)
- it is computer independent
- it can modify flexible molec.fragments, and accepts more sets in ATMOD
- it analyses all possible atoms sets obtained from the Patterson
  ( both for heavy atom structures as for Vector Search procedures )
- it includes a generation program for flexible molecular fragments
- and all is fully automatic: no control data needed.
- but... the old DIRDIF manual has not yet been fully updated:
  and... the lis1 and lis2 listing files give too much information.

This is a beta-version
----------------------

Please tell me all problems regarding all applications.
My e-mail: PTBeurskens@hetnet.nl with subject: DIRDIF ! 
           [ because of SPAM, I won't read email with other subjects !]

Thank you in advance,
Paul T. Beurskens

