Skip to content

Adding new annotation from ModelSeed #968

Closed

Description

@JHL-452b

Current behavior:

Considering that MEMOTE includes reaction and metabolite annotations from ModelSEED as part of its scoring, I have generated a list of matching reactions and metabolites between Human-GEM and ModelSEED based on the SMILES information of the metabolites.

A total of 4,391 metabolites and 2,680 reactions were successfully matched. The addition of these annotation details will enhance the score of Human-GEM in MEMOTE.

You can find more details in this All_Seed_mapping_mets.xlsx and All_Seed_mapping_rxns.xlsx.

Activity

  1. mihai-sysbio commented on Mar 20, 2025

    @mihai-sysbio
    Member

    I believe the annotation should end up in the corresponding TSV files.

  2. feiranl commented on Mar 21, 2025

    @feiranl
    Collaborator

    Thanks Mihail! Totally agree! Please proceed @JHL-452b by expanding the corresponding tsv files

  3. added
    wipA pull request for this issue is open
    fixed in developMerged into develop, will be in the next release
    and removed
    wipA pull request for this issue is open
    on Sep 24, 2026
  4. edkerk commented on Sep 27, 2026

    @edkerk
    Member

    #1105 corrects some of the ModelSEED ids that #1086 added. A check of each id against its ModelSEED equation or compound found that about 1-2 % point to a different reaction or compound. These ids are now corrected or removed. The other ids are unchanged.

    Reactions: id of a different reaction removed

    Reaction Removed id Why
    MAR03964 rxn09694 (kept rxn00062) proton-pumping variant on a plain ATP hydrolysis
    MAR03975, MAR03977, MAR03979, MAR06912, MAR07794 rxn42569 (kept rxn00001) same, on plain diphosphate hydrolysis
    MAR04207 rxn42210 (kept rxn00439) same
    MAR02530 rxn45259 18:2 elongation; the model reaction is 18:3 (the id stays on MAR02423)
    MAR00267, MAR02961 rxn33912 degenerate ATP + CoA reaction; it was matched through elaidoyl-CoA, which carried the ModelSEED id of elaidate
    MAR02031 rxn06311 plant reaction, reversed and unbalanced
    MAR08573 rxn41233 (kept rxn08936) celloheptaose reaction
    MAR08773 rxn19443 D-ribulose 1-phosphate aldolase; the model reaction uses D-xylulose 1-phosphate

    Reactions: proton coupling. #1086 keeps H+ when it crosses a membrane, so that proton-coupled and plain transporters stay apart. 11 transport ids did not follow that rule. 5 are replaced by the proton-coupled ModelSEED reaction: MAR05099 rxn09339 -> rxn05301, MAR05101 rxn08625 -> rxn09678, MAR05102 rxn08625 -> rxn13301, MAR05107 rxn08747 -> rxn05244, MAR05118 rxn05245 -> rxn05669. The other 6 are removed because ModelSEED has no reaction with the right coupling: MAR04964, MAR06327, MAR06381, MAR06382, MAR06604, and MAR06921 (complex I without proton pumping).

    Compounds

    Metabolite Before -> after Why
    MAM00188c cob(II)alamin (enzyme-bound) cpd12872 -> cpd27507 cpd12872 is the cob(I)alamin form
    MAM01224c 9,13-di-cis-retinoate cpd11177 -> removed cpd11177 is 9-cis-retinoate
    MAM02454c mannose-1-phosphate cpd00736 -> cpd00485 cpd00736 is a generic hexose 1-phosphate
    MAM02683c, MAM02330g paragloboside cpd02989 -> cpd03740 cpd02989 is Lc4Cer

    Related changes in other open PRs:

    MAR01444 and MAR08892 carry rxn08762 (bacterial Kdp). The id matched the model equation, but the model's H+/K+-ATPase left out H+ efflux. #1106 corrects the stoichiometry of MAR08892 and merges MAR01444 into it (see #1111); rxn08762 no longer matches and should be revisited there.

  5. added
    wipA pull request for this issue is open
    and removed
    wipA pull request for this issue is open
    on Sep 27, 2026
  6. added this to the 2.1.0 milestone on Sep 28, 2026
  7. edkerk commented on Oct 5, 2026

    @edkerk
    Member

    Fixed in Human 2.1.0.

  8. added
    featureSomething new: a reaction, pathway, dataset or function
    and removed
    fixed in developMerged into develop, will be in the next release
    on Oct 14, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    featureSomething new: a reaction, pathway, dataset or function

    Type

    No type

    Projects

    No projects

      Milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions