Add Support for DSSP v4.0

[cbalbin-bio]

• I hereby agree to dual licence this and any previous contributions under both
  the Biopython License Agreement AND the BSD 3-Clause License.

• I have read the CONTRIBUTING.rst file, have run pre-commit locally,
  and understand that AppVeyor and TravisCI will be used to confirm the Biopython unit
  tests and style checks pass with these changes.

• I have added my name to the alphabetical contributors listings in the files
  NEWS.rst and CONTRIB.rst as part of this pull request, am listed
  already, or do not wish to be listed. (This acknowledgement is optional.)

Closes #3763 #3433

This is a patch to the current DSSP parser to support DSSP v4.0 . The DSSP parser currently only handles up to v3.x.x. I think supporting v4.0 is important as that is what is install by default when using apt on Ubuntu 21.04 or greater. Currently it is not mentioned in the documentation that v4.0 is not supported and this could be undesirable to some users.

DSSP v4.0 shifts to a mmCIF format by default. You can output a DSSP v3.0 style report using the flag --output-format=dssp. While I am a big fan of mmCIFs gain in popularity. DSSP 4.0 s mmCIF out is missing some critical items such as accessibility; see comment.
Thus I think it is better to stick with the classic DSSP format output.

DSSP v4.0 also now uses auth_asym_id for the chain IDs instead of label_asym_id when ran on mmCIF files see comment. For older versions of DSSP, a mapping had to be done from the label to auth chain ID f85e4d9. This is now disabled when a DSSP v4.0 or higher executable is detected.

In order to enable the expected experience for the most users, whether using PDB or mmCIF format files, or DSSP v3 vs v4 I added a version check to the initializer of the DSSP class. This check uses code similar to that found the the DSSP unit test. If it is found that the executable has a version greater than or equal to '4.0.0' the flag --output-format=dssp is added to the dssp command. Also if the user is found to be using v4.0.0 or greater with an mmCIF file. The label to auth chain ID mapping is disabled as the results already contain the auth chain ID instead.

Thanks for your time,

Christian

[codecov]

Codecov Report

Merging #3775 (bbb114b) into master (c6da294) will increase coverage by 82.11%.
The diff coverage is 27.77%.

❗ Current head bbb114b differs from pull request most recent head b43f4b5. Consider uploading reports for the commit b43f4b5 to get more accurate results

@@             Coverage Diff             @@
##           master    #3775       +/-   ##
===========================================
+ Coverage        0   82.11%   +82.11%     
===========================================
  Files           0      298      +298     
  Lines           0    49955    +49955     
===========================================
+ Hits            0    41023    +41023     
- Misses          0     8932     +8932     

Impacted Files	Coverage Δ
Bio/PDB/DSSP.py	53.17% <27.77%> (ø)
Bio/SeqIO/__init__.py	84.66% <0.00%> (ø)
Bio/Align/Applications/_ClustalOmega.py	100.00% <0.00%> (ø)
Bio/ExPASy/ScanProsite.py	30.15% <0.00%> (ø)
Bio/SVDSuperimposer/__init__.py	86.95% <0.00%> (ø)
Bio/AlignIO/Interfaces.py	90.90% <0.00%> (ø)
Bio/Align/nexus.py	76.31% <0.00%> (ø)
Bio/phenotype/pm_fitting.py	12.28% <0.00%> (ø)
Bio/Blast/NCBIWWW.py	77.17% <0.00%> (ø)
Bio/GenBank/Record.py	91.96% <0.00%> (ø)
... and 289 more

Help us with your feedback. Take ten seconds to tell us how you rate us.

[rob-miller]

This is working well for me (once I read the docs on how to skip the file type test) and I support merging.

May I suggest changing make_dssp_dict() to use as_handle() instead of open():

from Bio.File import as_handle
...
def make_dssp_dict(filename):
    with as_handle(filename, "r") as handle:
        return _make_dssp_dict(handle)

This maintains current functionality and further enables the user to get the dssp output from their own subprocess if/when the one in this code fails again.

Also

    dssp : string
        The dssp executable (ie. the argument to subprocess)

might be better described as something like

    dssp : string
        The name of the dssp executable ('dssp' or 'mkdssp', ie. the argument to subprocess)

[rob-miller]

Just to note, I've run this on 6,145 mmCIF chains from a current Dunbrack PISCES nonredundant list, and the only issue is an alignment error on PDB code 3sqz chain A due to Cys 111 being a non-standard residue. The PDB format file for this chain is handled correctly.

[JoaoRodrigues]

Going back to this, thanks for the ping @rob-miller. I downloaded DSSP versions 3.x.x and 4.x.x (from their github repos) and installed them locally. I am having an issue with reading mmCIF files with version 4.x.x. The tests will pass on version 3.x.x but version 4.x.x produces an empty dssp_dict file. I am looking into it, but it might be related with issues with label_ and auth_ fields in the cif file.

[JoaoRodrigues]

Uh oh, sorry... my git pull origin master created this mess!