Pathway: oxidative phosphorylation (GO:0006119)

[ValWood]

Pathway: oxidative phosphorylation (GO:0006119)

Complete gene set https://www.pombase.org/term/GO:0006119

The model is missing the
proton-transporting ATP synthase complex (GO:0045259)/
proton motive force-driven mitochondrial ATP synthesis (GO:0042776) sub-part

because I don't have a COmplex Portal ID for proton-transporting ATP synthase complex (GO:0045259)
https://www.pombase.org/term/GO:0045259
(I could add it with the complex sub-parts)
See geneontology/go-ontology#29067

Model Link: http://noctua.geneontology.org/workbench/noctua-visual-pathway-editor/?model_id=gomodel%3A663d668500000596

Current Model:

GOCAM viz

GOCAM viz complexes uncollapsed

[ValWood]

If we agree this is the way to model this process, the fly pathway should be identical, except ndi1 will be substituted by complex I

[hattrill]

Building the Fly version - not connected anything or filled in all the gaps yet: http://noctua.geneontology.org/workbench/noctua-visual-pathway-editor/?model_id=gomodel%3A67086be200000314

[hattrill]

Added some connections - think will have to add H+ input to complex V and add Cytochrome c and electron transferring flavoproteins somewhere in the mix

[hattrill]

Adding this note: CHEBI has electron entry....CHEBI:10545

[ValWood]

Added some connections - think will have to add H+ input to complex V and add Cytochrome c and electron transferring flavoproteins somewhere in the mix

If you can add these parts I will copy them.

I will try to rationalize the assembly factors (I'll open a separate ticket)
we have a good start https://www.pombase.org/results/from/id/e5831c48-bd63-425d-92b1-c9dcd8aad95e
but there is some slightly wonky ontology ). I want to find out if these are really all assembly factors and, if so, what (i.e. chaperone) - some have turned out to be stabilizing the mRNA, so assembly is just a phenotype; this will take a while because I'm busy next week or so.

[hattrill]

1. Modified to show H+ output/input from complexes to complex V
2. Added cyt c
3. Very roughly added ETF complex and Etf-QO as place-holders - they mediate e- transfer from other mitochondrial pathways to ubiquinol for input for complex III. PMID: 33450351, but do not know what BP to add here as where boundary on oxphos chain is is unclear to me just now.

[ValWood]

Systematic ID	Gene name	Product description	Budding yeast orthologs	Human orthologs	Fly
SPAC27D7.06	etf1	electron transfer flavoprotein alpha subunit EtfA	AIM45	ETFA	wal
SPAC1805.02c	etf2	electron transfer flavoprotein beta subunit EtfB	CIR1	ETFB	Etfb
SPAC20G8.04c	cir2	mitochondrial electron transfer flavoprotein-ubiquinone oxidoreductase Cir2	CIR2	ETFDH	Etf-QO

[ValWood]

CHATGPT says

Yes, human ETFA, ETFB, and ETFDH are components of the mitochondrial electron transfer flavoprotein complex (ETF). They work together to facilitate the transfer of electrons from specific flavoprotein dehydrogenases to the respiratory chain via electron transfer flavoprotein
oxidoreductase (ETF-QO).

Breakdown:
ETFA (Electron Transfer Flavoprotein Alpha): The alpha subunit of ETF.
ETFB (Electron Transfer Flavoprotein Beta): The beta subunit of ETF.
ETFDH (Electron Transfer Flavoprotein Dehydrogenase): A flavoprotein that serves as an intermediary, receiving electrons from ETFA/ETFB and transferring them to the ubiquinone pool in the mitochondrial inner membrane.

so it seems that all 3 are involved...

Also,
These proteins collectively form part of the electron transfer flavoprotein system, which is crucial for fatty acid β-oxidation and the catabolism of some amino acids. The ETFA and ETFB subunits form the heterodimeric ETF protein, while ETFDH connects this system to the respiratory chain.

[ValWood]

I bunged them in my model so I don't forget , will connect them up later

[hattrill]

That's kind of what I did - you have reminded me of it now!

[ValWood]

Note to self, locate the ETF Regulatory Factor 1 (ETFRF1) (an LYR family member) in yeast
candidates
https://www.pombase.org/results/from/id/9e404030-051b-4123-8c76-44269b7db1cf

candidate from FoldSeek:
https://www.pombase.org/gene/SPCC2H8.04

[Antonialock]

For technical reasons, human has an additional complex I subunit (isoform of NDUFC2)
"NADH dehydrogenase [ubiquinone] 1 subunit C2, isoform 2, NDUFC2-KCTD14 readthrough transcript protein" NDUFC2-KCTD14.

I asked if I should merge this isoform into NDUFC2 but was told not to as "NDUFC2 & NDUFC2-KCTD14 are treated as 2 separate genes in resources like HGNC, Ensembl & NCBI Gene so it makes sense to keep the UniProt records separate too."

[ValWood]

Or even any evidence that it forms a functional protein (most readthrough RNAs would be degraded wouldn't they ?)

[Antonialock]

if I align the two products they look like this

NDUCR_HUMAN is the read-through

[ValWood]

But there is no KCTD14 here so why not just 2 isoforms

or not, just flag that one as a readthrough

sp|Q02827|NDUC2_BOVIN NADH dehydrogenase [ubiquinone] 1 subunit C2 OS=Bos taurus OX=9913 GN=NDUFC2 PE=1 SV=1
MMTGRQGRATFQFLPDEARSLPPPKLTDPRLAFVGFLGYCSGLIDNAIRRRPVLLAGLHR
QLLYITSFVFVGYYLLKRQDYMYAVRDHDMFSYIKSHPEDFPEKDKKTYGEVFEEFHPVR

The HPXR is clearly conserved so this is def the functional form.

Ensembl has tonnes of Readthrough transcripts that aren't represented in UniProt, so why this particular one?

[Antonialock]

I don't know! Maybe something historical? Since I was told I can't merge the UniProt records as "HGNC, Ensembl & NCBI Gene " treat them as separate records, I thought I should approach someone else to see if they would merge/oboslete. Perhaps HGNC?

[ValWood]

Yes, I would mail HGNC. As far as I can tell, there is no evidence for this existing as a real protein isoform. The C-term of the protein seems highly conserved, so this would be quite a severe mutant.

The AlphaFOld model confidence drops away too (because this isn't seen in any other species).

I guess this is one of the downsides of nanopore sequencing technology, it picks up these artefacts. It's a pity nobody has figured methods to screen them.

[Antonialock]

I compared complex I subunits from four resources: GO, wikipedia page, PMID:30030361 and structure of bovine complex 1.

thankfully good agreement, but I'll look at the few subunits that don't match up.

	30030361	wikipedia	GO	pdb:5LDW	comment
NDUFS8	x	x	x	x
NDUFAB1	x	x	x	x
NDUFA1	x	x	x	x
NDUFS4	x	x	x	x
NDUFB5	x	x	x	x
NDUFB3	x	x	x	x
NDUFC1	x	x	x	x
NDUFA2	x	x	x	x
NDUFS5	x	x	x	x
NDUFS7	x	x	x	x
NDUFS2	x	x	x	x
NDUFS6	x	x	x	x
NDUFB1	x	x	x	x
NDUFS3	x	x	x	x
NDUFB6	x	x	x	x
NDUFA3	x	x	x	x
NDUFB4	x	x	x	x
NDUFB8	x	x	x	x
NDUFB2	x	x	x	x
NDUFA7	x	x	x	x
NDUFC2	x	x	x	x
NDUFA10	x	x	x	x
NDUFB10	x	x	x	x
MT-ND1	x	x	x	x
MT-ND2	x	x	x	x
MT-ND3	x	x	x	x
MT-ND4L	x	x	x	x
MT-ND4	x	x	x	x
MT-ND5	x	x	x	x
MT-ND6	x	x	x	x
NDUFB7	x	x	x	x
NDUFS1	x	x	x	x
NDUFV1	x	x	x	x
NDUFA8	x	x	x	x
NDUFV3	x	x	x	x
NDUFA6	x	x	x	x
NDUFB11	x	x	x	x
NDUFA13	x	x	x	x
NDUFA12	x	x	x	x
NDUFB9	x	x	x	x
NDUFA5	x	x	x	x
NDUFV2	x	-	x	x
NDUFA9	-	x	x	x
WDR93	-	-	x
NDUFA11	-	x	x	x
NDUFA4	-	x	-		from paper: NDUFA4 was initially thought to be part of complex I [170], but later assigned to complex IV [110,117].
NDUFA4L2	-	x	-
NDUFC2-KCTD14	-	-	x

[ValWood]

nice!

[ValWood]

Actually, I don't think there are many readthroughs in UniProt human. Thre are a couple that look like they are multidomain proteins (and the 'readthrough" is conserved in mouse)

[Antonialock]

if respiratory cahin complex 1 has a number of catalytic subunits, and a number of accessory subunits, then could you remind me how to model that in noctua?

presumably just the catalytic subunits should be in the complex activity unit?

what happens to the other subunits?

[ValWood]

I am not sure how to do this.
This is part of the reason I requested geneontology/noctua#901
In your situation I would want to annotate some subunits with the activity, and some subunits with MF "unknown", or other activities if they are known.

@vanaukenk do you have a suggestion what to do here?
Maybe we could discuss further in the New Year?