~6 incorrect annotations

[oxinabox]

I was resolving the yeastGEM against MetaNetX v4.4 and found some collisions.
Most that I found I think were mistakes in MetaNetX, but some I think are mistakes in the identifiers in the yeastGEM.
I list the collisions below headed by the metanetx name and URL, and below that the info from all colliding molecules in the GEM, I bold the section I think is wrong or dubious:

• alpha-D-ribose 5-phosphate (https://www.metanetx.org/chem_info/MNXM1104056)
  • s_1408 (ribose-5-phosphate [c])
    • sbo: SBO:0000247
    • bigg.metabolite: r5p
    • chebi: CHEBI:18189
    • kegg.compound: C03736
    • metanetx.chemical: MNXM15900
  • s_3768 (D-ribofuranose 5-phosphate [c])
    • sbo: SBO:0000247
    • chebi: CHEBI:78346
    • kegg.compound: C00117
    • metanetx.chemical: MNXM722712
  • s_3852 (aldehydo-D-ribose 5-phosphate [c])
    • sbo: SBO:0000247
    • chebi: CHEBI:58273
    • kegg.compound: C00117
    • metanetx.chemical: MNXM722712
• taurocholate (https://www.metanetx.org/chem_info/MNXM1104296)
  • s_1473 (taurocholic acid [c])
    • sbo: SBO:0000247
    • bigg.metabolite: tchola
    • chebi: CHEBI:36257
    • kegg.compound: C05122
    • metanetx.chemical: MNXM2288
  • s_4267 (taurocholate [c])
    • sbo: SBO:0000247
    • bigg.metabolite: tchola
    • biocyc: CPD-3743
    • chebi: CHEBI:28865
    • kegg.compound: C05122
    • metanetx.chemical: MNXM1104296
• 6-(N-acetyl-alpha-D-glucosaminyl)-1-phosphatidyl-1D-myo-inositol (https://www.metanetx.org/chem_info/MNXM1104777)
  • s_0331 (6-(N-acetyl-alpha-D-glucosaminyl)-1-phosphatidyl-1D-myo-inositol [er])
    • sbo: SBO:0000247
    • bigg.metabolite: acgpail_hs
    • chebi: CHEBI:57265
    • kegg.compound: C01288
    • metanetx.chemical: MNXM999
  • s_3879 (G00143 [er])
    • sbo: SBO:0000247
    • chebi: CHEBI:57265
    • metanetx.chemical: MNXM92477
• alpha-D-galactose (https://www.metanetx.org/chem_info/MNXM1108175)
  • s_0558 (D-galactose [c])
    • sbo: SBO:0000247
    • bigg.metabolite: gal
    • chebi: CHEBI:12936
    • kegg.compound: C00124
    • metanetx.chemical: MNXM390
  • s_3862 (alpha-D-Galactose [c])
    • sbo: SBO:0000247
    • bigg.metabolite: a_gal__D
    • chebi: CHEBI:28061
    • kegg.compound: C00984
    • metanetx.chemical: MNXM390
• raffinose (https://www.metanetx.org/chem_info/MNXM1108339)
  • s_3717 (protein [c])
    • sbo: SBO:0000649
    • bigg.metabolite: protein
    • kegg.compound: C00492
    • metanetx.chemical: MNXM621
  • s_4193 (raffinose [c])
    • sbo: SBO:0000247
    • bigg.metabolite: raffin
    • chebi: CHEBI:16634
    • kegg.compound: C00492
    • metanetx.chemical: MNXM621
• succinate (https://www.metanetx.org/chem_info/MNXM25)
  • s_1460 (succinate [m])
    • sbo: SBO:0000247
    • bigg.metabolite: succ
    • chebi: CHEBI:30031
    • kegg.compound: C00042
    • metanetx.chemical: MNXM25
  • s_4313 (ITP [m])
    • sbo: SBO:0000247
    • bigg.metabolite: succ
    • biocyc: SUC
    • chebi: CHEBI:61402
    • kegg.compound: C00081
    • metanetx.chemical: MNXM25
    • pubchem.compound: 160419
    • seed.compound: cpd00036

I am not sure though i have bolded the right bits as errors.
Especially for taurocholate which kegg does indeed list as an alias for taurocholic acid

The total affected ID are:

['s_1408', 's_3768', 's_3852', 's_1473', 's_4267', 's_0331', 's_3879', 's_0558', 's_3862', 's_3717', 's_4193', 's_1460', 's_4313']

but as you can see the problems are probably only in about half of them

I actually ran this against v9.0.0, but i have checked the release notes and the diffs of PRs that were merged and I don't see these sections as changed. I might have missed though.

Related issues:

• fix-update MNX IDs according to MNX database latest release 4.0 #245
• duplicate metabolites identified #338

[oxinabox]

• D-xylulose (https://www.metanetx.org/chem_info/MNXM731281)
  • s_0580 (D-xylulose [c])
    • sbo: SBO:0000247
    • bigg.metabolite: xylu__D
    • chebi: CHEBI:17140
    • kegg.compound: C00310
    • metanetx.chemical: MNXM597
  • s_4180 (L-xylulose [c])
    • sbo: SBO:0000247
    • bigg.metabolite: xylu__L
    • chebi: CHEBI:17140
    • kegg.compound: C00312
    • metanetx.chemical: MNXM597

I am sure this one s_4180 (L-xylulose [c]) is wrong as it's CHEBI is supposed to be:
CHEBI:17399

[oxinabox]

• L-methionine (S)-S-oxide (https://www.metanetx.org/chem_info/MNXM1106045)
  • s_3807 (L-methionine (S)-S-oxide [c])
    • sbo: SBO:0000247
    • chebi: CHEBI:49031
    • kegg.compound: C15999
  • s_3837 (L-Methionine S-oxide [c])
    • sbo: SBO:0000247
    • bigg.metabolite: metsox_S__L
    • chebi: CHEBI:17016
    • kegg.compound: C02989

s_3807 (L-methionine (S)-S-oxide [c]) should have bigg.metabolite: metsox_S__L
s_3837 (L-Methionine S-oxide [c]) should have bigg.metabolite:metox

[edkerk]

Thank you for reporting this. Indeed please clearly indicate what change should be applied to the model to fix the incorrect annotation. Like:

s_3807 (L-methionine (S)-S-oxide [c]) should have bigg.metabolite: metsox_S__L
s_3837 (L-Methionine S-oxide [c]) should have bigg.metabolite:metox

[oxinabox]

I'm not a biologist, its hard for me to work out what the right answers for various things are.
I mostly just know that the things I have bolded in my top post are wrong.
Or in some cases that they can not all be correct.
For my purposes it has been enough to just drop them, as I would rather no annotation than incorrect ones.

I do have a full script now that does drop everything I am dubious of, that was blocking me from resolving against metanetx v4.4, and confirmed it is working on yeastGEM v9.0.2 now
And it is now resolving most molecules, or declearing it can't match them (about 10% mostly ones that have no annotations except SBO).
Which is good enough for my purposes right now.

Some of these dubiously have no right answer apparently bigg actually does use the same ID for taurocholate and (taurocholic acid.
Which like are actually different things, in that they have different InCHi, and one is a a base and the other is an acid
So like perhaps it is correct to attach that ID to both.