rsref.py parameter_file [-r] [-nostat]
-r if the refinement is to be run on a remote computer
-nostat cancels the default window of refinement statistics
rsref.py is a Python script that runs the programs (Selectcoord, Refinement by CNS, Reexpand) in a batch of real-space refinement. It controls the various programs coordinately using parameters read from parameter_file which specifies, for example, whether optimization is by conjugate gradient or simulated annealing, the residues to be refined etc..
Simulated annealing refinement should be used if large changes are required (to get out of a local minimum). Conjugate gradient is appropriate for conservative refinement near the optimum.
RSdir – the full path of directory which contains the RSRef2000 installation, e.g. ser/c/users/RSRef2000. This directory should already contain subdirectories pscripts, omacros etc. if correctly installed.
Unix operating systems:
RSdir would normally be set automatically on log-in with a command such as setenv RSdir ser/c/users/RSRef2000 in .cshrc for c-shell users.
Windows operating systems (front-end workstation):
RSdir is set in Start 4 Control Panel 4 System 4 Environment.
Environmental variables for CNS. Normally, this is done on log-in with a command such as source /full_path_for_CNS/cns_solve_env in .cshrc for c-shell users.
See the CNS for full details.
REM_MACHINE – the account and computer name where RSRef2000 is to be run remotely, in the format account@computer.
REM_DIR - the full directory path containing the RSRef2000 installation on the remote computer.
These would normally be set on log-in the same way as RSdir (see above) under both unix and windows. So, the following example commands in .cshrc would be appropriate for c-shell users: setenv REM_MACHINE firstname.lastname@example.org; setenv REM_DIR leu/users/RSRef2000.
These are created in the RSRef_Temporary_Fls subdirectory of user's account on the remote and local computer where refinement is run. The files are purged following normal completion of the refinement.