Running MODELLER

To run MODELLER with the script file model-default.py above, do the following:

1. Open a command line prompt:
   • On Linux/Unix: ssh to the machine, or open an xterm or GNOME Terminal.
   • On Windows: Click on the `Modeller' link on your Start Menu. This will give you a Windows Command Prompt, set up for you to run MODELLER.
   • On Mac OS X: ssh to the machine, or open the Terminal application.

2. Change to the directory containing the script and alignment files you created earlier, using the 'cd' command.

3. Run MODELLER itself by typing the following at the command prompt:

   python model-default.py > model-default.log

(Note that if you don't have Python installed on your machine, you can run the last step by typing 'mod9.14 model-default.py' instead.)

A number of intermediary files are created as the program proceeds. After about 10 seconds on a modern PC, the final 1fdx model is written to file 1fdx.B99990001.pdb. Examine the model-default.log file for information about the run. In particular, one should always check the output of the alignment.check() command, which you can find by searching for `check_a'. Also, check for warning and error messages by searching for `W>' and `E>', respectively. There should be no error messages; most often, there are some warning messages that can usually be ignored.

(Note: if you have Python installed on your system, you may be able to use MODELLER as an ordinary Python module. This works automatically on Mac OS X, Windows with any version of Python between 2.3 and 3.4, or using the RPM install on Linux with any version of Python between 2.3 and 3.4. On other systems, you may need to set the PYTHONPATH and LD_LIBRARY_PATH variables, or create symlinks to the relevant directories, so that your system can find the MODELLER Python modules and dynamic libraries, respectively.)