updates disvis tutorial page

[joaomcteixeira]

This PR corrects DisVis tutorial regarding the swapping of the structure names (PRE5 and PUP2). This change does not affect the overall message of the tutorial, but it improves the detailed results.

The problem: PRE5.pdb and PUP2.pdb file names are swapped in the original data.

What I did to correct the tutorial:

1. When downloading the disvis-tutorial.tgzfile and inspecting the PDB files in detail we see that the PRE5.pdb contains the sequence of PUP2 (UniProtKB: Q9UT97) and PUP2.pdb contains the sequence of PRE5 (UniProtKB: O14250).
2. I swapped name of PDB files
3. I changed the new PRE5.pdb chain B to A and new PUP2.pdb chain A to B. In this way the pdbs are in agreement with the chains in restraints.txt. I used:
   1. pdb_chain -A PRE5.pdb > pre5.pdb
   2. pdb_chain -B PUP2.pdb > pup2.pdb
   3. mv pub2.pdb PUB2.pdb
   4. mv pre5.pdb PRE5.pdb
   5. new files can be downloaded here: https://transfer.sh/jwefzD/new_structures.zip
4. open them in chimera first PRE5 than PUP2 (as said in the tutorial)
5. Sent the first tutorial run with these settings: https://wenmr.science.uu.nl/disvis/run/kwO_3K3%2BbmqJ
6. after the run we see that the Number of accessible complexes 0 for 9 restraints, while it was 0 for 7 restraints in the original version
7. False positives, with the corrected version we find:
   1. A164-B49 (2.38) -> original version was 1.45
   2. A49-B188 (0.48) -> original version was 1.23
8. identified false positives as restraint 5: "A 164 CA B 49 CA 0 23". Before the false positive was ambiguous. (restraint 3 (0.85), 5, and 9), where 3 is in the interface when visualised in the corrected structures.
9. Question: "Using the different descriptions of the sections we provided above together with the information on the results page of your run, what are likely false positive restraints according to DisVis?"
   1. here i answered only restraint 5, but maybe A49-B188 can also be considered even with z-score 0.48 because it is the second highest
10. the N threshold is 3 to 4 when visualising accessible_interaction_space.mrc.
11. removed only restraint 5 to generate a new restraints_filtered.txt. But maybe we could remove also restraint 9.
12. swapped lines in accessible_res_70.list
13. did the complete scanning also with occupancy analysis: submitted second job: https://wenmr.science.uu.nl/disvis/run/kwO_3K4aijxJ
14. the number of restraints in the Interaction Analysis table is 8 instead of 7. Identified residues above 0.5:
    1. PRE5 (new): 3, 5, 8, 11, 53, 54, 55,56, 58, 60, 79, 82 , 122, 123, 124
    2. PUP2 (new): 13, 15, 17, 18, 19, 126, 127, 128, 129, 130, 131, 162, 164, 165, 180, 184,
15. Following the description of 5L5A we can see that the original input structured were really swapped as the original tutorial says 5L5A chain D corresponds to PRE5 and this is still true even after the PDBs names are changed.
16. Made some chimera sessions to show the distances, and the interactions. They can be downloaded here: https://transfer.sh/BKsQPu/chimera_sessions.zip
17. We can see in the corrected example there is no "alone Leucine" as in the original tutorial.

If you agree with this, I need help in defining which files go in the disvis_tutorial.tgz, namely the second_run_with_mrc.py, and the results folder. I haven't investigated yet if this branches to other tutorial pages.

[amjjbonvin]

The disvis-tutorial.zip is also part of the repo and should be updated before merging. And similarly the output example pages of DISVIS must be updated on the server side.

And finally - this will also impact the HADDOCK-MS-crosslinks tutorial which builds on the results from this one.

[joaomcteixeira]

Okay, I wanted to have your approval first, I will look into updating what's missing.