Generate fragment file on my computer

Using make_fragments.pl is the older way to make fragments. It's a little cumbersome to use, primarily because there's a fair amount of configuration that needs to be done before it can run. (If you open the script, at the top there are paths to various external programs and databases which need to be set.) But once you actually get it working, it does work appropriately. (The trick is setting things up in the first place.)

The newer version of the fragment picker uses a compiled program in the rosetta/main/source/bin directory instead of the Perl script. (See https://www.rosettacommons.org/docs/latest/application_documentation/utilities/app-fragment-picker for the documentation.) This is the protocol used in the demos/public/fragment_picker demo.

The actual running of the fragment picker is much more straight-forward in the newer case. HOWEVER,  much of the reason for this is that it offloads a bunch of the complexity to preparation steps. You still need the external programs to make secondary structure predictions, etc. except with the new method you run them manually, as opposed to setting the path to the programs in the script and having things run automatically.

But the BestFragmentsProtocol in the demo directory should be rather simple, as only need an (appropriately formatted) psipred secondary structure prediction file. If you have more information (for example, homolog alignments, chemical shifts, other experimental data, etc.) you can have more complex protocols which give slightly more accurate fragment predictions, but for a first pass of fragment picking, the BestFragmentsProtocol should give you something that works with Rosetta.