modified on 27 January 2012 at 17:59 ••• 29,930 views

Geometry load

From NWChem

Jump to: navigation, search


   [ LOAD [format xyz||pdb]  [frame <int frame>] \
          [select [not] \
               [name <string atomname>] \
               [rname <string residue-name>]
               [id  <int atom-id>|<int range atom-id1:atom-id2> ... ]
               [resi <int residue-id>|<int range residue-id1:residue-id2> ... ]
   <string filename> ]

LOAD directive allows users to load Cartesian coordinates from external pdb or xyz files with the name <filename>. This directive works in addition to the explicit Cartesian coordinate declaration and can be repeated and mixed with the latter. This allows for complex coordinate assemblies where some coordinates are loaded from external files and some specified explicitly in the input file. The ordering of coordinates in the final geometry will follow the order in which LOAD statements and explicit coordinates are specified.

  • The actual file from which coordinates will be loaded is presumed to be located in the run directory (the same place where input file resides). Its name cannot coincide with any of the keywords in the LOAD statement. To keep things simple it is advised to specify it either at the beginning or the end of the LOAD directive.
  • format xyz || pdb - specifies format of the input file. The only formats that are supported at this point are pdb and xyz. Either one can contain multiple structures, which can be selected using the frame directive. Note that in case of PDB file multiple structures are expected to be separated by END keyword. If the format directive is not provided the format will be inferred from file extension - .xyz for xyz files and .pdb for pdb files
  • frame <int frame> - specifies which frame/structure to load from multiple structure xyz or pdb files. In the absence of this directive the 1st frame/structure will be loaded
  • select ... - this directive allows to selectively load parts of the geoemtry in the file within aspecified frame. Selections can be based on atom index (id), atom name (name), and for pdb files can also include residue index (resi) and residue name (rname). Atom and residue name selection are based on an exact single name match. Atom and residue index allow both multiple single number and multiple range selection. For example
 select id 2 4:6 9

will result in the selection of atom id's 2 4 5 6 9.

Multiple selection criteria are always combined as AND selections. For example

  select name O  id 2:4

will select atoms that are named O and whose id/index is between 2 and 4. Each selection criteria can be inverted by prepending not keyword. For example

  select not name O  id 2:4

will select all atoms that are not named O and whose id/index is between 2 and 4.