Metabolism terms - add missing logical definitions

[paolaroncaglia]

Though logical definitions (aka ‘cross-products’) have been added broadly to the ontology, a few stray terms have escaped the effort. Making a note here as my task to look at the ‘metabolic process’ branch and add the missing defs.

(Note for self: previous work was recorded in a Jira ticket
https://www.ebi.ac.uk/panda/jira/browse/GO-110
now closed, but may wish to look into its more recent comments.)

[paolaroncaglia]

Note for self - David H and Peter noted that it'd be good to do some of this similarly to glycolysis; David to add comments/examples here. Thanks @ukemi

[paolaroncaglia]

Note for self: this is low priority, but I'll leave it assigned to me for the time being, in case we decide I should (be the one to) address it.

[paolaroncaglia]

Hi @pgaudet and @cmungall
Here, too, I noticed the discussion on old tickets in the minutes of the latest ontology call. Obviously at this point I'm not going to be working on this ticket any time soon. If you think that it's still relevant, could you please assign it to someone else? Else, I guess it can be closed according to your recent discussion. Thanks.

[pgaudet]

(not yet finished)

Here are terms for which I don't know how to add a logical definition

• GO:0036001 'de novo' pyridoxal 5'-phosphate biosynthetic process
• 'GO:0019681 acetyl-CoA assimilation pathway'
• GO:0019676 ammonia assimilation cycle
• ergothioneine biosynthesis from histidine via N-alpha,N-alpha,N-alpha-trimethyl-L-histidine' see Query: ergothioneine biosynthesis from histidine via N-alpha,N-alpha,N-alpha-trimethyl-L-histidine' versus ergothioneine biosynthesis from N-alpha,N-alpha,N-alpha-trimethyl-L-histidine' #15256
• GO:0030633 D-alanine family amino acid catabolic process 0 annotations; is this needed?

I think these are MFs and should be merged into their parent:

• GO:0019465 aspartate transamidation no annotations; isn't this a MF?
• GO:0019802 cyclization of glutamine involved in intein-mediated protein splicing -> Merge into intein-mediated protein splicing

---

Does this seem OK ?
tyrosine biosynthetic process, by oxidation of phenylalanine
'biosynthetic process'
and ('has input' some phenylalanine)
and ('has output' some tyrosine)
and ('has part' some 'oxidation-reduction process')

lysine biosynthetic process via diaminopimelate, dehydrogenase pathway
'biosynthetic process'
and ('has intermediate' some '2,6-diaminopimelate(2-)')
and ('has part' some 'diaminopimelate dehydrogenase activity')
and ('has output' some lysinate)

lysine biosynthetic process via diaminopimelate, diaminopimelate-aminotransferase pathway
'biosynthetic process'
and ('has intermediate' some '2,6-diaminopimelate(2-)')
and ('has part' some 'diaminopimelate epimerase activity')
and ('has part' some 'L,L-diaminopimelate aminotransferase activity')
and ('has output' some lysinate)

transsulfuration
'metabolic process'
and ('has input' some homocysteine)
and ('has output' some cysteine)
and ('has intermediate' some cystathionine)

[pgaudet]

Many terms not yet done, including:

'D-alanine family amino acid catabolic process': 0 annotations
'D-alanine family amino acid metabolic process'
'D-alanine family amino acid biosynthetic process'
-> perhaps this should be CHEBI 26463 - pyruvate family amino acid?
An L-α-amino acid which is biosynthesised from pyruvate (i.e. alanine, valine, and leucine).

'anaerobic L-alanine catabolic process'
'L-alanine catabolic process, by transamination'
'L-alanine oxidation to D-lactate and ammonia'
'L-alanine oxidation to pyruvate via D-alanine'

'anaerobic glycine catabolic process'
'glycine decarboxylation via glycine cleavage system'

'lysine biosynthetic process via aminoadipic acid'

''de novo' L-methionine biosynthetic process'
'L-methionine biosynthetic process from L-homoserine via cystathionine'
''de novo' NAD biosynthetic process from aspartate'

'L-methionine biosynthetic process from homoserine via O-acetyl-L-homoserine and cystathionine'
'L-methionine biosynthetic process from homoserine via O-phospho-L-homoserine and cystathionine'
'L-methionine biosynthetic process from homoserine via O-succinyl-L-homoserine and cystathionine'

'L-methionine salvage from methionine sulphoxide'
'L-methionine salvage from methylthioadenosine'
'L-methionine salvage from S-adenosylmethionine'

'aspartate transamidation'

'L-lysine catabolic process to glutarate, by acetylation'
'L-lysine catabolic process to acetyl-CoA via 5-aminopentanamide'
'L-lysine catabolic process to acetyl-CoA via L-pipecolate'
'L-lysine catabolic process to acetyl-CoA via saccharopine'
'L-lysine catabolic process using lysine 6-aminotransferase'
'L-lysine catabolic process to glutarate, by acetylation'
'lysine biosynthetic process via aminoadipic acid'

'S-adenosylmethionine cycle'

'anaerobic phenylalanine oxidation'
'arginine deiminase pathway'

'anaerobic glutamate catabolic process'
'glutamate catabolic process via 2-hydroxyglutarate'
'glutamate catabolic process via mesaconate and citramalate'
'glutamate catabolic process via L-citramalate'
'glutamate catabolic process to succinate via 2-oxoglutarate-dependent GABA-transaminase activity'

'histidine catabolic process to imidazol-5-yl-lactate'

'anaerobic L-alanine catabolic process'
'L-alanine catabolic process, by transamination'
'L-alanine oxidation to D-lactate and ammonia'
'L-alanine oxidation to pyruvate via D-alanine'

'L-cysteine catabolic process via cystine, using cysteine transaminase'
-> cysteine transaminase -> not in GO
....
to be continued !

[cmungall]

Can we make a push on this? Maybe I can spend an hour at the hackathon going through these? I'm very fast at adding these.

D-alanine family: can we make a chebi ticket for this?

I think this is a priority for the quality of the metabolic branch. Without axioms and tracing through to the actual chemical entities there are ambiguities as it's just words. With ambiguities we have lower quality, and inconsistent annotations. Also xrefs to pathway databases are suspect. With the chemical axioms in place we can automatically align to pathway databases in a robust fashion. For an example of a problem area see #17435

[pgaudet]

Or if you have time before.
Is it possible to get the list of all terms with no LD ?

I think the new relations 'has primary input/output' should help for terms such as 'glutamate catabolic process via 2-hydroxyglutarate'.

As well, a number of terms should be probably be obsoleted, such as 'anaerobic catabolism of pairs of amino acids' (this is clearly a function).

Pascale