GPRs for Mitochondrial 3-Hydroxyacyl-CoA Oxidation Reactions

[Devlin-Moyer]

Current behavior:

MAR03269 and MAR03380 are both associated with HSD17B4 and EHHADH and both represent oxidation of a 3-hydroxyacyl-CoA to a 3-oxoacyl-CoA in mitochondria:

ID	Reaction	Current GPR
MAR03269	NAD+ + 3(S)-3-Hydroxydodecen-(5Z)-Oyl Coenzyme A <=> H+ + NADH + 3-Oxolaur-Cis-5-Enoyl Coenzyme A	HSD17B10 or HADHA or EHHADH or HSD17B4 or HADHB or HADH
MAR03380	NAD+ + (3S)-3-Hydroxy-Cis-8-Tetradecenoyl Coenzyme A <=> H+ + NADH + 3-Oxo-Cis-8-Tetradecenoyl Coenzyme A	HSD17B10 or EHHADH or HSD17B4 or HADH

There are a number of other mitochondrial reactions associated with both of these genes, but they all have other issues that I address in #738, so I'm not mentioning them here.

Expected behavior:

While HSD17B4 and EHHADH can both catalyze the oxidation of 3-hydroxyacyl-CoAs, they are both localized to peroxisomes and not mitochondria; the mitochondrial enzymes that can catalyze this reaction are HADHA (ENSG00000084754), HADH (ENSG00000138796), and HSD17B10 (ENSG00000072506).

Uniprot mentions that HADH can only act on substrates with carbon chain lengths of up to 10, so I don't think it should be associated with either of these reactions, since the reactants of MAR03269 and MAR03380 have chains with 12 and 14 carbons, respectively.

The best paper I could find on the chain length preferences of HADHA says it can handle chain lengths between 6 and 16, and I haven't been able to find any papers that clarify what chain lengths HSD17B10 can handle, so I think both HADHA and HSD17B10 are fine to keep in the GPRs of these reactions.

Proposed changes:

• Change the GPR of MAR03269 to ENSG00000084754 or ENSG00000072506
• Change the GPR of MAR03380 to ENSG00000084754 or ENSG00000072506

[haowang-bioinfo]

agree to assign ENSG00000084754 or ENSG00000072506 as GPR of MAR03269 and MAR03380

for example, the comp-mismatch of HSD17B4 and EHHADH can be found and reported by the code presented in #746

[haowang-bioinfo]

fixed by #760