Repository navigation
fix: Human-GEM errors that the animal GEMs inherit - #1168
Merged
Merged
Conversation
…ity diff A gene outside a cell line's context model has no knockout effect there, but the diff skipped every line where the gene was in the context model on one side only. A change that adds a gene to a GPR, or takes it out, was therefore reported as unchanged. Such lines are now compared, with the gene non-essential on the side where it is absent. The CSV gains a model_membership column for genes new to or removed from the model. A gene missing from a matrix that is older than its model is not compared, and matrices without a shared cell line are refused. This also corrects the grouping of the report: genes removed from or added to the model get rows of their own instead of sharing a row with genes that stay, where the membership label of one gene read as applying to the whole row.
TRDMT1 methylates tRNA, and DNMT3L is catalytically inactive; DNMT1 and DNMT3B do not need each other. Both genes leave the model.
…1156) MMAA is the G-protein chaperone of MMUT. MAR08614 (cobalamin import into mitochondria) has no known transporter; MAR08615 and MAR08616 are MMAB's. MAR08616 is made irreversible. MMAA leaves the model.
PEDS1 (plasmanylethanolamine desaturase, MAR07601), COQ4 (MAR06906, EC 4.1.1.130), PRUNE1 (triphosphate hydrolysis, MAR05363), FADS2 (24:4n-6 desaturation, MAR02403, renamed), and MMACHC or MTRR for aquacobalamin reduction (MAR08613, now NADPH, with a note). PEDS1, COQ4 and MMACHC are new genes.
#1159, #875) 274 mitochondrial and peroxisomal beta-oxidation reactions: isoenzymes become alternatives (or), chain-length and substrate-class specificity follows the literature, and MYO5B, a myosin, leaves every thiolase GPR and the model. The bile-acid transport MAR01694 loses its thiolase GPR and EC codes, the 2-methyl-branched oxidase steps MAR03498 and MAR03512 get ACOX2 or ACOX3, and MAR03460 (ECH1) gets EC 5.3.3.21.
…antoinase (#1147) Removes MAR06611, MAR06610 and MAR06609, deprecates 5-hydroxyisourate, and drops EC 3.1.1.4 from the allantoin transport MAR10003.
…1150) Adds 4-hydroxymandelate synthase (HPDL, MAR20218), (S)-4-hydroxymandelate and a gene-free lumped step to 4-hydroxybenzoate (MAR20219). Removes the tyrosine -> 4-coumarate steps, which have no human enzymes, and the cytosolic copy of the coumarate route; the mitochondrial route from dietary 4-coumarate stays, gene-free. CellFie task 178 (ubiquinone from tyrosine) follows the new route.
) MAR07597 lumped alkyl-DHAP reduction with an acylation by the free fatty acid pool, without the cost of acyl-CoA. It is replaced by MAR20220 on the acyl-CoA pool; the reduction is MAR03354. The phosphatase MAR03383 becomes irreversible: run backwards it phosphorylated 1-alkyl-2-acetylglycerol with free phosphate, a cycle that MAR20220 would otherwise close.
Removes MAR02314, MAR03437, MAR03438, MAR03439 and MAR03483, corrects the isomers in MAR00968, MAR02397, MAR02399, MAR02497, MAR02499, MAR03415, MAR03417, MAR03440, MAR03474 and MAR03479, takes MAM00902 back from the deprecated list and adds MAM20093; MAR03440 consumes the proton of its twin MAR02395, and MAM00853x gets the n-6 structure its name describes. The eight ER elongation steps get the name, EC, GPR, direction and subsystem of their enzymes (ELOVL2, HSD17B12, HACD1-4, TECR). Rhea and MetaNetX ids follow the corrected reactions.
MAR08615 becomes 2 cob(II)alamin + NADPH => 2 cob(I)alamin + NADP+ + H+, the reduction of MMAB-bound cob(II)alamin within EC 2.5.1.17; NADPH, as the FDXR-FDX1 system that supplies the electrons in vitro. It was written as cob(II)alamin + NADH <=> cob(I)alamin + H+ + NAD+, which does not balance.
Without H+ exchange, 27 of these 47 tasks passed only through charge-unbalanced reactions. The ER tasks keep their closed proton balance.
…olites (#1118, #1140) MAR00022 takes lipoic acid instead of [protein]-N6-(lipoyl)lysine, which nothing makes, so MAR00021 is feasible again; it is renamed from "[protein]-N6-(lipoyl)lysine reduction" to "Cofactors and vitamins pool for MAR00021". The vitamin A, D and E derivative pools and their pool reactions are removed.
(S)-3-hydroxybutanoyl-CoA[x] (MAM00173x) occurs in MAR03104 only, so the peroxisomal dehydrogenase could never carry flux. Both are deprecated.
Contributor
Model quality report❌ 1 regression(s) vs
Regression(s):
Improved: ✅ unchanged · ✨ improved vs Reaction directions: the alarm and the warning catch only directions that thermodynamics makes very unlikely. A reaction without an alarm or warning is not thereby shown to have the right reversibility. Full workflow run · this comment is edited as results come in |
This was referenced Oct 20, 2026
Member
Author
|
Gene essentiality was run on this pull request via GH Actions. Gene essentiality: effect of this changeChecked 53 gene(s) directly affected plus 2230 more that share a metabolite with one of them. A gene counts as changed below only if the flip agrees in direction across at least 3 of the 5 cell-line models; fewer than that is grouped as likely noise from ftINIT's run-to-run variability instead.
Full detail: gene-essential_summary.md. ✅ unchanged · ✨ correct vs Hart 2015 · ❌ wrong vs Hart 2015 · ℹ️ not comparable to Hart 2015 The unscoped, every-gene matrix behind the table above is committed to gene-essential.csv.
|
The animal GEMs number their own reactions and metabolites in MAR15000-MAR19999 and MAM15000-MAM19999. qcModelChecks.py now fails if the model, its annotation tables or the deprecated lists use that block, and lists the identifiers in qc_reserved_identifiers.csv. The gate is silent while clean: buildReport.py shows its row only when it finds something.
MMACHC reduces aquacobalamin with glutathione and only without oxygen (https://doi.org/10.1074/jbc.M114.567339), and no measurement shows it with NADPH. `MAR08613` keeps MTRR, which does reduce it with NADPH (https://doi.org/10.1073/pnas.0603694103). MMACHC leaves the model, as no other reaction uses it. #1149
…ve reaction (#1149) MAR06906 now converts 3-methoxy-4-hydroxy-5-all-trans-decaprenylbenzoate to 2-methoxy-6-all trans-decaprenyl-2-methoxy-1,4-benzoquinol with O2 and NADPH under COQ4. MAR06907, which assigned COQ6 to the C1 hydroxylation, is deprecated.
Member
Author
|
The gene-essentiality run on this pull request lists five genes as wrong against Hart 2015 (FADS2, H6PD, MMAB, PRUNE1, HPDL) and four as mixed (AKR1A1, GNPAT, SLC22A5, COQ4). None of these is a regression that should block the merge:
The essential calls of FADS2, PRUNE1, MMAB, HPDL and H6PD are false positives that follow from biomass content and from an unknown alternative CoQ route. They are listed under "Expected effect" in #1167 and do not indicate an error in the changed GPRs. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #1155, closes #1156, closes #1147, closes #1150, closes #1151, closes #1152, closes #1159, closes #1167. They are grouped because each one affects the animal GEMs built from Human-GEM (#421); #1167 lists what is left for later.
Part of #1149 (
MAR01506, LRAT, is left), #1130 (the GPRs of the third and fourth β-oxidation steps), #1120 (MAR07597) and #1118 (the cofactor pool ofMAR00021, #1140).Based on develop 87fa3a1, which includes #1162. The trimethyllysine-release change of #1157 is therefore no longer part of this PR.
Reactions
Removed (deprecated)
MAR06609,MAR06610,MAR06611MAR00040,MAR01724,MAR11357,MAR06783,MAR06784,MAR06785,MAR06786MAR07597MAR02314,MAR03437,MAR03438,MAR03439,MAR03483MAR00023,MAR00033,MAR00034MAR00022: deprecatedMAR06907MAR06906, deprecatedMAR03104MAM00173xsince #929: deprecated, withMAM00173xAdded
MAR20218MAR20219MAR20220Equations, bounds, names and ECs
MAR08615MAR08616MAR08613MAR00022MAR00968,MAR02397,MAR02399,MAR02497,MAR02499,MAR03415,MAR03417,MAR03440,MAR03474,MAR03479MAR03412,MAR03415,MAR03417,MAR03418,MAR03440,MAR03442,MAR03474,MAR03479MAR03417,MAR03418,MAR03474,MAR03479reversibleMAR02403MAR06906MAR06907), EC 4.1.1.130MAR03460MAR01694MAR10003MAR03383GPRs (295 reactions)
MAR04072,MAR08640,MAR08641MAR08614;MAR08615,MAR08616MAR07601;MAR06906;MAR05363;MAR02403;MAR08613HSD17B4 or EHHADH,ACAA1 or SCP2,(HADHA and HADHB) or HADH or HSD17B10,ACOX2 or ACOX3Totals: 21 reactions removed, 3 added, 295 GPRs changed, 14 equations corrected.
Metabolites
MAM00173xMAR03104only)MAM00982c,MAM00996c,MAM01004m,MAM01104cMAM00853c,MAM00853rMAM00902MAM03139c,MAM03140c,MAM03143cMAM00902cMAM00902r,MAM20093c,MAM20093rMAM20092mGenes
Tasks and code
qcModelChecks.py: reaction and metabolite numbers 15000 to 19999 (MAR15000-MAR19999,MAM15000-MAM19999) belong to the animal GEMs, which number their own reactions and metabolites there. The Model QC workflow fails when the model, its annotation tables or the deprecated lists use one, and lists it inqc_reserved_identifiers.csv.model/README.mddescribes the numbering and the reserved block, and a comment ingenerateAnimalGEM.pynames the block.metabolicTasks_Essential.txt: the IC, BS and SU tasks may exchange H+ (Essential tasks allow no proton exchange, so 28 of them pass only through charge-unbalanced reactions #1152).metabolicTasks_CellfieConsensus.txt: task 178 follows the new CoQ head-group route (The CoQ head group is made through a plant 4-coumarate route instead of HPDL #1150).diffGeneEssentiality.py: a gene that is in one side's context model only is scored in that line; matrices older than their model are not compared; no shared cell line raises an error; amodel_membershipcolumn. Develop's markers for genes that enter or leave the model are kept. Genes that leave or enter the model now get report rows of their own, so their label no longer reads as if it covered the other genes in the row.Maintainers: after merge, credit new contributors by commenting
@all-contributors please add @username for <type>on this PR or the linked issue (see CONTRIBUTING).