Open JHL-452b opened 2 months ago
Due to the difficulty in distinguishing enzyme promiscuity in lipid metabolism, I recommend for this issue we only remove ENSG00000162365 from MAR00934, as it is clearly stated that ENSG00000197408 shows no activity towards arachidonic acid and prostaglandin A1. @edkerk what do you think?
Indeed, these curations involve a lot of Cytochrome P450s with not uncommon wide substrate specificity.
Remove ENSG00000162365 from MAR00934
(see Feiran), and ENSG00000197408 from MAR00932
(involves in epoxidation, making EET, not HETE).
Not suggested by the AI, but ENSG00000186377 should be removed from all reactions in the model, except for MAR00954
, MAR00955
and MAR00956
.
If anything, for Cytochrome P450s (who can generally be identified by there gene abbreviation starting with CYP), it might be more reasonable to curate per gene: has its protein been assayed for activity --> only keep the gene in that reaction --> remove the gene from all other reactions. But this involves manual curation that is outside the scope of the current approach.
Current behavior:
27 genes are discovered to be uncorrelated with their corresponding reactions in
Arachidonic acid metabolism