Open JHL-452b opened 1 month ago
The AI is right that ACOT4 (ENSG00000177465
) shouldn't be in the GPR of MAR00249
, since it doesn't catalyze that reaction, but ACOT2 (ENSG00000119673
) should also be removed from the GPR for the same reason. This paper found that ACSBG2 (ENSG00000130377
) showed no activity towards stearate (C18:0), so it should also be removed from the GPR of MAR00249
.
It's also right about MYO5B (ENSG00000167306
) not belonging in the GPRs of MAR03315
or MAR03093
, but both ACAA1 (ENSG00000060971
) and SCP2 (ENSG00000116171
) can independently catalyze these reactions, so the GPRs for both reactions should be ENSG00000060971 or ENSG00000116171
)
It's also right about ECI1 (ENSG00000167969
) not belonging in the GPR of MAR03067
, but there are other problems with that GPR:
ENSG00000198721
) catalyzes the same reaction as ECI1, which is not what's happening in MAR03067
ENSG00000084754
) is mitochondrial, not peroxisomalENSG00000133835
) can also catalyze this reaction in peroxisomes, so it should be ORed with EHHADH (ENSG00000113790
)So the GPR of MAR03067
should just be EHHADH or HSD17B4 (ENSG00000113790 or ENSG00000133835
). All of these problems also apply to MAR03091
, so its GPR should also be ENSG00000113790 or ENSG00000133835
.
MAR03102
and MAR03103
should just be removed from the model, since beta-oxidation of acyl-CoAs with fewer than ~8 carbons happens in mitochondria and not peroxisomes, and the mitochondrial versions of those reactions are already represented by MAR03163
and MAR03164
, respectively (sources: 1, 2).
The AI is right about ACOX1 (ENSG00000161533
) not being able to catalyze MAR03506
, but it could be replaced with ACOX2 (ENSG00000168306
), since both ACOX2 and ACOX3 (ENSG00000087008
) are capable of oxidizing branched acyl-CoAs (source).
Nothing is wrong with the GPR of MAR03338
; both ACOX1 (ENSG00000161533
) and ACOX3 (ENSG00000087008
) are absolutely capable of catalyzing that reaction (sources: 1, 2). I suspect that it is once again confused by the fact that we decided to make all of the peroxisomal acyl-CoA oxidation reactions reduce FAD to FADH2 instead of O2 to H2O2 in #639.
Thanks for your investigation and the detailed feedback. @Devlin-Moyer I totally agree with all your comments after checking it out. In summary,
MAR00249
should be ENSG00000068366 or ENSG00000103740 or ENSG00000123983 or ENSG00000140284 or ENSG00000151726 or ENSG00000164398 or ENSG00000197142
after removing ENSG00000177465
, ENSG00000119673
and ENSG00000130377
.MAR03315
should be ENSG00000060971 or ENSG00000116171
.MAR03093
should be ENSG00000060971 or ENSG00000116171
.MAR03067
should be ENSG00000113790 or ENSG00000133835
.MAR03091
should be ENSG00000113790 or ENSG00000133835
.MAR03506
should be ENSG00000087008 or ENSG00000168306
.MAR03338
should be ENSG00000087008 or ENSG00000161533
without any change.MAR03102
and MAR03103
should be removed from the model.LGTM!
Current behavior:
21 genes are discovered to be uncorrelated with their corresponding reactions in
Linoleate metabolism
,Fatty acid elongation
andBeta oxidation of even-chain fatty acids
.