Duplicate IDs for same mets

[feiranl]

Description of the issue:

mets	metsNoComp	metBiGGID	metKEGGID	metHMDBID	metChEBIID	metPubChemID	metHepatoNET1ID	metRecon3DID	metMetaNetXID	metHMR2ID	metRetired
MAM00270c	MAM00270	wharachd	C14748	empty	CHEBI:34306	5283157	empty	wharachd	MNXM22451	m00270c	m00270c
MAM00270r	MAM00270	wharachd	C14748	empty	CHEBI:34306	5283157	empty	wharachd	MNXM22451	m00270r	m00270r
MAM00591c	MAM00591	empty	C14748	empty	CHEBI:34306	empty	HC02179	empty	MNXM6760	m00591c	m00591c
MAM00591e	MAM00591	empty	C14748	empty	CHEBI:34306	empty	HC02179	empty	MNXM6760	m00591s	m00591s
MAM00270e	MAM00270	wharachd	C14748	empty	CHEBI:34306	5283157	empty	wharachd	MNXM22451	empty	m00270s

[haowang-bioinfo]

this shows there are two sets ids CHEBI:34306 (KEGG:C14748) in [c] and [e], which set should be removed?

[feiranl]

We will keep the metabolite MAM00270 , since it is related to 12 rxns (7 related to MAM00591)
Duplicate rxns caused by this merge will also be removed(the right column).

rxn with MAM00270	GPR	rxn with MAM00591	GPR
MAR00934	Many genes	MAR00941	ENSG00000162365 or ENSG00000187048
MAR02396	ENSG00000134538	MAR06142	ENSG00000134538
MAR02398	ENSG00000134538	MAR06143	ENSG00000134538
MAR02402	ENSG00000174640	MAR06220	ENSG00000174640
MAR02404	ENSG00000184999	MAR06252	Blank
MAR02400	ENSG00000134538	MAR06144	ENSG00000134538
MAR10348	Blank	MAR10025	Blank

[haowang-bioinfo]

Duplicate rxns caused by this merge will also be removed(the right column).

then how to deal with the GPRs after removal?

[feiranl]

These two metabolites are mapped to the same metabolite according to the MetaNetX

mets	metsNoComp	metBiGGID	metKEGGID	metHMDBID	metChEBIID	metPubChemID	metLipidMapsID	metEHMNID	metHepatoNET1ID	metRecon3DID	metMetaNetXID	metHMR2ID	metRetired
MAM00077x	MAM00077	CE2416						CE2416			MNXM165274	m00077p	m00077p
MAM02766x	MAM02766	prist		HMDB0000795	CHEBI:51340	123929	LMPR0104010022	CE0932		prist	MNXM3342;MNXM7698	m02766p	m02766p

I think we should remove the metabolite MAM00077x, as it only relate to 2 rxns, while the other one is associated with 13 rxns from all compartments.

****Associated rxns for MAM00077x
MAR03488
Action: remove, for the wrong annotation of EC number with this rxn. EC1.14.11.18 should be related to AKG, while this rxn is more like 1.2.1.3.

MAR03489 is a duplicate of MAR03387 with the duplicate met MAM02766x , so it is okay to remove. CAUTION! GPR conflict.

[feiranl]

Duplicate rxns caused by this merge will also be removed(the right column).

then how to deal with the GPRs after removal?

I am not sure about this. In this case, there is one rxn which has conflict(I have updated the table). Do you have any suggestions?

[haowang-bioinfo]

Do you have any suggestions?

According to above table with updated reactions and GPRs, all GPRs for the left column can be retained except MAR00934, whose new GPR might be changed by merging GPR of MAR00941? So that there will be no information loss.

[haowang-bioinfo]

I think we should remove the metabolite MAM00077x, as it only relate to 2 rxns, while the other one is associated with 13 rxns from all compartments.

agree

MAR03489 is a duplicate of MAR03387 with the duplicate met MAM02766x. CAUTION! GPR conflict.

a tentative solution is merging genes from both if no more evidence, e.g. use the version from MAR03489 whose GPR includes all genes from GPR of MAR03387

[haowang-bioinfo]

fixed in #534