Curations that would enable Escher visualizations

[djinnome]

Hi folks,

This is an attempt to convert my email from May 11th, 2019 to an issue that can be tracked.

There is a jupyter notebook that contains all the suggested fixes. I will attempt to break each suggestion into a separate comment below.

Hi Jeremy,

Some good news – Ed Kerkhoven from Chalmers has been working with Christian Lieven from DTU to add BiGG IDs to their Yarrowia model (https://github.com/ChristianLieven/Yarrowia_lipolytica_W29-GEM/tree/bigg_ids). They regularly post updates and we can link to their latest release for our paper.

Ed and Christian plan to publish an updated model in another paper that is associated with some data that was produced at EMSL awhile ago and I think there is some interest is collaborating with you on visualization as they develop their new model. You can follow-up with Ed directly to assess how you could collaborate.

Cheers,
Scott

Hi Scott,

That is fantastic news!

Ed, what kind of visualizations you are interested in?

Sincerely,

Jeremy

Hi Jeremy,

You're probably aware of Escher (https://escher.github.io/)? It would be nice if the model works with that.

Some other curations that we're planning to include are BIGG anntations, Yali1 gene identifiers, curate with reactions from iYLI647 and iMK735, SLIMEr formulation of lipid metabolism (https://bmcsystbiol.biomedcentral.com/articles/10.1186/s12918-018-0673-8), do you have some additional ideas after working with the model?

Best, Ed

[djinnome]

iMM904 central carbon metabolism reactions missing in YALI4

• FRUK: f16_c and fdp_c have no other reactions.
• ALCD2x_copy1 iMM904 had two copies of this reaction for no good reason. ACLD2ir is the correct reaction in yarrowia. Recommend replacing the reaction in the map
• ETOHt: y001762 (etoh_c[c] <=> etoh_e[e]) is the equivalent reaction in YALI4. Is there any reason not to rename this reaction to ETOHt?
• ACt2r: Closest reaction is y001106 but no proton symport and very different GPR
• ACtm: No ability to directly transport acetate from mitochondria to cytosol without converting to Acetyl-CoA in Yali4
• MALt2r: y001901 converts L-malate from extracellular to cytosol without proton symport
• CITt2r: y001686 transports citrate from extracellular to cytoplasm without proton symport
• ACONTm: Explicit representation of citrate -> cis-aconitate -> isocitrate in Yali4. Recommend modifying escher map to accomodate
• SUCD1m: YALI4 does not contain a SUCD1m reaction, but all homologs to yeast SUCD1m genes are present in Yali4. Recommend adding this reaction to YALI4
• FRDm: YALI4 does not contain a FRDm reaction, but homologs to yeast FRDm genes are present in Yali4. Recommend adding reaction to YALI4
• FUMt2r: Closest reaction is y001796 fum_e[e] <=> fum_c[c] with no proton symport. Recommend altering YALI4 to accommodate proton symport
• SUCOASm: Identical in stoichiometry to y001022: adp_m[m] + pi_m[m] + succoa_m[m] --> atp_m[m] + coa_m[m] + succ_m[m]
  Yeast is reversible, but Yarrowia reaction is not. Recommend renaming y001022 to SUCOASm
• AKGDbm: Identical stoichiometry and reaction direction to y000831. Recommend renaming
• GCC2cm_copy1: GCC2cm_copy1 is duplicated in iMM904 for no good reason. Identical stoichiometry to y000505. Reaction is reversible in Yeast and irreversible in yarrowia. Recommend renaming to GCC2cm
• AKGDam: Nearly identical stoichiometry to y000832 with the exception of sdhlam_m, which has one less hydrogen in its chemical formula than HC00695_m. Note that HC00695 is a universal BIGG metabolite with an instance in the iCHOv1 model.
  Recommend aligning with Yeast sdhlam_m metabolite. See missing sdhlam_m discussion below
• CO2t: Identical stoichiometry to y001697. Recommend renaming

[edkerk]

Thanks @djinnome. It is not always entirely clear what curation you propose. For instance about ALCD2x_copy1. However, some are just changes of a reaction ID to the one used in the Escher map? Others seem to be changes in the reaction equation, e.g. changing fumarate to proton symport. How sure are you that these reactions in iMM904 (and then specifically in that Escher map) is correct? The Yarrowia lipolytica model is based on yeast-GEM 8, which is a more up-to-date version than iMM904, so is it not possible that those reactions have since been curated? It seems a bit abitrary to make the model fit a map of an old model?