Inconsistent IDs for FMN and FMNH2

[Devlin-Moyer]

Current behavior:

I was starting to work on implementing the changes I proposed in #609, and I noticed inconsistencies in the metabolite IDs for FMN and FMNH2 metabolites in different compartments:

ID	Name	Compartment	Formula	Charge
MAM01828c	FMN	c	C17H19N4O9P	-2
MAM01828x	FMN	x	C17H19N4O9P	-2
MAM01828e	FMN	e	C17H19N4O9P	-2
MAM20020r	FMN	r	C17H18N4O9P	-2
MAM20023i	FMN	i	C17H18N4O9P	-2
MAM20019r	FMNH2	r	C17H21N4O9P	-2
MAM20022i	FMNH2	i	C17H21N4O9P	-2

(These are all of the metabolites currently in Human-GEM that have "FMN" in their names. Also I bolded the number of hydrogens in the empricial formulas just to make it clear that that's the only part that changes)

Expected feature/value/output:

• [X] All of the FMN metabolites should have the same IDs except for the compartment suffix, so MAM20020r should be MAM01828r and MAM20023i should be MAM01828i.
• [X] Same for the FMNH2 metabolites: MAM20022i should be MAM20019i.
• [X] The empirical formulas for all FMN metabolites should be the same. According to ChEBI, PubChem, and KEGG, the formula for (neutral) FMN should be C17H21N4O9P, and since these represent the -2 ion of FMN, the formulas for MAM20020r and MAM20023i should be changed to C17H19N4O9P.
• [X] All of the FMN metabolites should be associated with the same KEGG ID, so MAM20020r and MAM20023i should both be associated with C00061.
• [X] Relatedly, neither of the FMNH2 metabolites are currently associated with a KEGG ID, so they can both be associated with C01847.

[haowang-bioinfo]

thanks for the scrutiny, go ahead to implementation please

[Devlin-Moyer]

FYI I just edited my proposed changes a bit because I realized the empirical formulas I looked up were for neutral FMN and FMNH2, but these metabolites represent their -2 ions since their phosphate groups deprotonate in water at physiological pHs.

[haowang-bioinfo]

yes, KEGG often picks neutral state for metabolites - in fact the protonation state under pH 7.3 is favoured

[haowang-bioinfo]

let's keep this issue open until the charge-balancing fixed