[x] Make mouse model: gomodel:6516135700000380, gomodel:6516135700000961, gomodel:6516135700001077
[x] Check reactants of R-HSA-71249 . According to PMID:11802770 the output is 4-N-trimethylaminobutyraldehyde (not in Chebi). They look like they are similar compounds. C7H16NO2+ versus C7H16NO https://pubchem.ncbi.nlm.nih.gov/compound/4-Trimethylammoniobutanoate#section=2D-Structure. The reactants look correct per Figure 3 of https://europepmc.org/article/ppr/ppr727037 . Also, the current output is acceptable to RHEA as input for the following reaction (Rhea17986). Ultra-weedy question: why are this and the next item on this list heme issues rather than carnitine biosynthesis #187 ones? The precedingEvent annotation in Reactome (4th item on this checklist) explains this.
[x] PMID: 20306513 suggests that THA1 may be the mouse enzyme that catalyzes this reaction, not SHMT1. https://europepmc.org/article/ppr/ppr727037 But in humans it is a pseudogene! I blasted it against the Candida one and it was not a significant hit. Curoiuser and curiouser.
Human THA1, unlike mouse, IS a pseudogene - see https://europepmc.org/article/ppr/ppr727037 for argument that human SHMT1 has the role in carnitine biosynthesis that Tha1 has for mice (don't think they checked for other functions of SHMT1 / Tha1 aside from carnetine biosynthesis - recent evolutionary divergence to add to our list along with purine catabolism and vitamin C synthesis.)
[x] Reaction "SHMT1 tetramer cleaves HTMLYS to yield TEABL and Gly" R-HSA-71249 is getting pulled into the heme biosynthesis pathway because it is manually annotated as a preceding event to "ALAS condenses SUCC-CoA and Gly to form dALA" (R-HSA-189442), but this is a manual curation error in Reactome, now fixed. Glycine is an output of 71429 and an input of 189442, but the first reaction generates cytosolic glycine and the second requires mitochondrial glycine. There is no cytosol-to-mitochondrial matrix reaction annotated here and as far as I know there is nothing special about the glycine produced by 71429 that makes it a preferred source for 189442. I deleted the preceding-following relationship, so 71249 should no longer be causally linked to heme biosynthesis.
[x] Check Reactome reaction charge and mass balance by conforming Reactome inputs and outputs to RHEA. Add RTHEA reaction names to Reactome reaction instances.
[x] Might want to add PMID:12474143 as a human reference for R-HSA-2995334. Done
[ ] editorial / scope question: is reaction 190141 "ALAD binds to Pb2+ " out scope for GO-CAM? This is a feature of lead poisoning, not normal biology. And if it is out of scope, could a non-NULL disease attribute on the reaction, added as part of the heme biosynthesis revision, work to get it filtered out of the GO-CAM build, in the same way that "drug" reactions are filtered? @dustine32
Cleanup notes: Heme notes.docx