refactor clustfcc and it's tests to use the proper path

[rvhonorato]

Re-organize clustfcc tests and make sure the module is using the proper paths - this is needed for #1013

[AnnaKravchenko]

I run a simple workflow:

[topoaa]
[clustfcc]
plot_matrix = true
[caprieval]

And got the following error:

[2024-09-11 16:30:55,365 __init__ INFO] [topoaa] took 4 minutes and 40 seconds
[2024-09-11 16:30:55,377 __init__ INFO] Running [clustfcc] module
[2024-09-11 16:30:55,377 __init__ INFO] Calculating contacts
[2024-09-11 16:30:55,377 libutil INFO] Selected 8 cores to process 55 jobs, with 8 maximum available cores.
[2024-09-11 16:30:55,377 libparallel INFO] Using 8 cores
[2024-09-11 16:30:58,375 libparallel INFO] 55 tasks finished
[2024-09-11 16:30:58,375 __init__ INFO] Calculating the FCC matrix
[2024-09-11 16:30:58,425 libutil ERROR] [Errno 2] No such file or directory: '1_clustfcc/fcc.matrix'
Traceback (most recent call last):
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/libs/libutil.py", line 335, in log_error_and_exit
    yield
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/clis/cli.py", line 192, in main
    workflow.run()
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/libs/libworkflow.py", line 43, in run
    step.execute()
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/libs/libworkflow.py", line 162, in execute
    self.module.run()  # type: ignore
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/modules/__init__.py", line 242, in run
    self._run()
  File "/Users/Kravc006/postdoc/tools/haddock3/src/haddock/modules/analysis/clustfcc/__init__.py", line 125, in _run
    with open(fcc_matrix_f, "w") as fh:
FileNotFoundError: [Errno 2] No such file or directory: '1_clustfcc/fcc.matrix'
[2024-09-11 16:30:58,428 libutil ERROR] [Errno 2] No such file or directory: '1_clustfcc/fcc.matrix'

The same workflow + the main branch runs smoothly

[rvhonorato]

Thanks @AnnaKravchenko, apparently there was no change necessary to the path of the clustfcc module, somehow it saves things to the proper path already 🔮

Running this test.toml

run_dir = "tt"
molecules =  [
    "data/e2aP_1F3G.pdb",
    "data/hpr_ensemble.pdb"
    ]
[topoaa]
[rigidbody]
sampling=10
[clustfcc]
plot_matrix=true
[caprieval]

Got to finish it with no issues and all the files in place:

$ ls -ltrh examples/docking-protein-protein/tt/2_clustfcc
total 68K
-rw-r--r-- 1 rodrigo rodrigo  316 Sep 12 12:38 params.cfg
-rw-r--r-- 1 rodrigo rodrigo  684 Sep 12 12:38 fcc.matrix
-rw-r--r-- 1 rodrigo rodrigo 1.9K Sep 12 12:38 clustfcc.txt
-rw-r--r-- 1 rodrigo rodrigo  310 Sep 12 12:38 clustfcc.tsv
-rw-r--r-- 1 rodrigo rodrigo  31K Sep 12 12:38 io.json
-rw-r--r-- 1 rodrigo rodrigo  11K Sep 12 12:38 fcc_matrix.html
-rw-r--r-- 1 rodrigo rodrigo  359 Sep 12 12:38 con.tgz
-rw-r--r-- 1 rodrigo rodrigo  100 Sep 12 12:38 cluster.out.gz

[AnnaKravchenko]

Is it normal that within that simple workflow ([topoaa]->[clustfcc]->[caprieval]) my second-best model is different? While the best model remains the same.
E.g.
run1 new branch: irmsd of the 2nd model 18.949
run1 main branch: irmsd of the 2nd model 34.830
run2 main branch: irmsd of the 2nd model 39.923
No reference structure given.

Is it because I did not actually score the models? And is it normal in general?

[amjjbonvin]

This is a very weird workflow as there is basically no HADDOCK score calculated at the topoaa stage. I.e. the ranking might be random. I would rather take as a test for example one from the docking-protein-protein example.

[AnnaKravchenko]

Indeed, it was the absence of a scoring module. With emscoring the results are the same