Appendix B - PDB to '.ins' Format Conversion with PDBINS

Appendix B - PDB to '.ins' Format Conversion with PDBINS


The auxiliary program PDBINS reads a PDB format file and writes a '.ins.' file for SHELXL-93. There are no command line options and the program runs interactively, asking the user to supply missing information. It is assumed that the PDB file conforms to the specifications in the PDB documentation 'Atomic Coordinate and Bibliographic Entry Format Description' of Feb. 1992. Atom lists from XPLOR and some other programs can be read, but if the conventions for transforming from orthogonal to crystallographic coordinates do not correspond to those in the above document, then appropriate SCALEn records must precede the ATOM or HETATM records. Note that SHELXL-93, unlike XPLOR, does not allow the same residue number to be used with different residue classes, and that the SHELXL-93 residue classes must begin with a letter and the SHELXL-93 residue numbers must be pure numbers and may not contain non-digits. XPLOR usually requires each solvent molecule to be defined as a separate residue, whereas for SHELXL-93 the solvent may either be treated in this way or assigned to the (default) residue number zero. PDBINS may renumber the residues if the current numbering scheme would be inconvenient for SHELXL-93. If there is more than one chain (or molecule) in the asymmetric unit, different residue numbers are required and there should be a gap of one or more residue numbers between different chains (otherwise there are problems with C_- and N_+ etc.).

PDBINS reads restraints and other standard instructions from a residue dictionary file 'shelxl.dic'. Users are encouraged to use this file as a model for their own dictionary files (which should be given different names). If the protein contains disordered or non-standard residues, some editing of the resulting '.ins.' file will be required before SHELXL-93 can be run. The order of atoms in each residue is irrelevant for SHELXL-93, but for PDBINS there are advantages in putting the atom named 'N' (the peptide nitrogen) first in each residue (as is normal practice). Some restraints may be missing if non-standard residue or atom names are used; note that PDBINS expects the C-terminal oxygen to be called OXT and not to be put in a separate residue. PDBINS converts OT to OXT for the C-terminus and CD to CD1 in isoleucine in accordance with PDB rules. In addition, PDBINS generates restraints for disulfide bridges linking residues CYS (or CSS). PDBINS uses the first character of the atom name as the element name, and recognizes two character element names if they start one column to the left, in accordance with PDB rules.

SHELXL-93 may be run with Friedel opposites NOT averaged (in which case no MERG instruction is needed; this is the correct option when significant anomalous scatterers e.g. iron (using CuKa) are present) or using MERG 4 to average Friedel opposites and set the dispersion terms f" to zero. Note that XPLOR and many other macromolecular programs, unlike SHELXL-93, require a reflection list in which Friedel opposites have already been merged.

PDBINS creates a '.pdl' file which gives residue names and compositions and other useful information; this file should be printed out and retained for reference during the refinement.

PDBINS is written in essentially machine independent FORTRAN-77. Before compiling it the comments in the source should be consulted and the following three changes may be required: (a) it is advisable to open the .ins and .pdl files with CARRIAGECONTROL='LIST' for VAX/VMS systems, (b) the only format statement in subroutine GETANS may be changed to end with ',$' to tidy up the console output for VAX/VMS and some other systems and (c) the variable CR may be set to CHAR(13) in the first executable statement in the main program if the program is to run on a UNIX machine which shares a disk with MSDOS machines using NFS network software. This enables DOS programs such as the Siemens SHELXTL-PC system to read the files directly. For pure UNIX AND VMS systems CR should be set to CHAR(32).


Ahead to C - Tables Production from '.cif' and '.fcf' Files using CIFTAB

Back to A - Absorption Corrections with SHELXA-93

Back to Appendices

Back to Table of Contents