Automated Scripts

ARP/wARP is a command file driven package. These command files, although simple in principle, turned out to be rather error prone. Many problems of previous releases were due to these kind of errors. This distribution has been complemented with a set of easy-to-use set-up scripts that we hope will help avoid mistakes and will generally improve the overall user-friendliness of the programs.

THE USER IS STRONGLY ADVISED TO RUN ARP/WARP USING THE AUTOMATED SCRIPTS PROVIDED WITH THE PACKAGE.

Of course, you may still generate your own command files - see the section on 'I hate automated scripts'. In fact, for the moment this might be a better choice for solvent building, especially towards the end of refinement where you might want to keep very good control of REFMAC parameters adjusting geometry weights, etc. In the next version we hope to include a graphical interface giving you full access to more REFMAC parameters rather than the basic set you can control now from the arp_warp_setup.sh script and the warp.par parameters file.

The current automated scripts are:

• arp_warp_setup.sh This is the basic script for setting everything up. Answering simple questions like which MTZ file, which labels to use, how many models, how many processors, etc. allow a general parameter file, warp.par, to be created which will be used for all subsequent jobs.

• arp_molrep.sh This is the script to run if you start from a reasonable Molecular Replacement solution (application #1). This script will use the model for providing geometrical restraints to the refinement. If your model is really bad, but you have data extending to at least 2.0 Å resolution, we suggest to use the warpNtrace.sh script (see below), which will auto-build a new model and will keep on auto-building, thus providing a significantly higher radius of convergence.

• warp.sh This is the script for applications #2 and #3. It allows to distribute jobs over a network, a multi-processor or a stand-alone machine. It also performs multiple model averaging if it is requested during setup. The script does not require any external programs (other than CCP4).

• warpNtrace.sh This script is for auto-tracing and model building (application #4). In some cases it can be considered to be used also for application #1, instead of arp_molrep.sh. Please note that this is an $\alpha$-version and only SGI executables are provided at present. In brief, the script reads either a map obtained after a job of warp.sh or a free-atoms model and does iterative auto-tracing and refinement. It can practically deliver a complete auto-trace of the model even with 2.0 Å data.

• side_dock.sh Takes a (partial) model auto-traced with warpNtrace.sh and docks the fragments into the protein sequence.

• arp_solvent.sh The good old solvent building is now automated to even greater extent. If you need to have a high degree of customisation of the REFMAC part of the job, we recommend to run the old-style command files, at least for the moment. Also, if you prefer to use SHELXL while water building, you would again have to use the old-style files, as provided in the examples/ directory. See the section 'I hate automated scripts' and the chapter on keywords.