Removed partial codon warning when testing amino acid sequences

[kaskales]

Only changed the testing amino acid sequences so the warning isn't triggered. Amino acid sequences don't need to be divisible by 3, only codon sequences.

Trimmed three testing protein_seqs by one amino acid to prevent irrelevant partial codon warning from appearing when testing TranslationError of amino acid Seq() object.

• 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 #...

[codecov]

Codecov Report

Merging #3307 into master will decrease coverage by 1.08%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #3307      +/-   ##
==========================================
- Coverage   83.98%   82.90%   -1.09%     
==========================================
  Files         318      316       -2     
  Lines       51661    51511     -150     
==========================================
- Hits        43389    42705     -684     
- Misses       8272     8806     +534     

Impacted Files	Coverage Δ
Bio/phenotype/pm_fitting.py	12.28% <0.00%> (-75.44%)	⬇️
Bio/PDB/mmtf/__init__.py	28.57% <0.00%> (-52.39%)	⬇️
Bio/codonalign/codonseq.py	44.16% <0.00%> (-44.31%)	⬇️
Bio/Phylo/_utils.py	28.57% <0.00%> (-35.72%)	⬇️
BioSQL/DBUtils.py	56.16% <0.00%> (-19.18%)	⬇️
BioSQL/BioSeqDatabase.py	68.61% <0.00%> (-9.31%)	⬇️
Bio/motifs/jaspar/db.py	3.08% <0.00%> (-5.41%)	⬇️
Bio/phenotype/phen_micro.py	83.81% <0.00%> (-2.26%)	⬇️
Bio/Phylo/BaseTree.py	88.93% <0.00%> (-0.21%)	⬇️
Bio/Seq.py	96.81% <0.00%> (-0.18%)	⬇️
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 63ec6b5...3f59651. Read the comment docs.

[peterjc]

I don't follow from the commit or pull request why you are changing this. If anything, it makes me wonder if there is a mistake in the tests?

[kaskales]

There is no mistake in the test that I see. I was just curious and running the test_seq tests when I noticed a warning,

BiopythonWarning: Partial codon, len(sequence) not a multiple of three. Explicitly trim the sequence or add trailing N before translation. This may become an error in future.
BiopythonWarning,

that was being triggered by the test_translation_on_proteins test while assertRaises(TranslationError) was being caught.

I just thought it might make the test pass cleaner if the warning wasn't being triggered, since it seems to be aimed for codon sequences and not about the amino acids being tested. Shortening the test amino acid sequences doesn't change any of the other tests.

[MarkusPiotrowski]

From a quick inspection it looks to me the reason that the test_translation_on_proteins passes in the first place is because a TranslationError is thrown for the wrong reason, i.e. that a 'codon' of three amino acids is not found in the table.

biopython/Bio/Seq.py

Lines 2407 to 2410 in 63ec6b5

	else:
	raise CodonTable.TranslationError(
	f"Codon '{codon}' is invalid"
	) from None

Since we cannot distinguish between nucleic and protein sequences in the Seq object any longer, the test which presumably has tested before that a protein cannot be translated, has become obsolete?

[peterjc]

Thanks @kaskales, test_translation_on_proteins was the scope I wanted clarified. Your change makes sense to me.

@MarkusPiotrowski Since the Seq object no longer distinguishes nucleotide versus protein, a protein/peptide "AAA" can be translated just fine but "MED" fails as it does not make sense as a codon. The test motivation still makes sense to me.

Perhaps the length not a multiple of three warning should happen later in the translation (allowing the exceptions to trigger first). Also perhaps if using to_stop=True the warning should be skipped too (if an in-frame stop is found)?

[MarkusPiotrowski]

@MarkusPiotrowski Since the Seq object no longer distinguishes nucleotide versus protein, a protein/peptide "AAA" can be translated just fine but "MED" fails as it does not make sense as a codon. The test motivation still makes sense to me.

Exactly. Because protein sequences may be translatable, this test, described as "Check translation fails on a protein." doesn't make sense any longer. If the sequence was "ATG" or "MAS", the test wouldn't fail.
There is another test, test_translation_of_invalid_codons, which could cover the "MED" case.

[peterjc]

OK, if test_translation_of_invalid_codons covers it we can drop test_translation_of_invalid_codons 👍

[kaskales]

Should I delete the test_translation_on_proteins, and push it to this PR, or make a new one for this?

[peterjc]

Whichever is easier for you.

[kaskales]

Are there any additional changes I should make for this PR?

[peterjc]

Over to @MarkusPiotrowski for second opinion (and merge)?

[peterjc]

This was resolved another way in #3446, so belated closing this now. Sorry @kaskales for not following up on this sooner.