GO term changes related to Fatty acid synthesis

[gocentral]

After reviewing some fatty acid synthase genes, I have several suggestions for changing related GO terms.

1. The definition of the following GO terms, GO 47451, GO 4317 and GO 47450 should be clarified by adding additional information that is available as a comment for the GO terms’ corresponding EC number at IUBMB and on the ExPASy website. Each of these GO terms correspond to a reaction that has an EC number. The GO term (and EC enzyme) is named after a specific substrate, but the comment specifies that the activity is actually specific to a range of substrates. I have added the corresponding comment for each of the GO terms below.

GO:0004317 3-hydroxypalmitoyl-[acyl-carrier-protein] dehydratase activity (EC 4.2.1.61)

Comment Specific for 3-hydroxyacyl-[acyl-carrier-protein] derivatives C(12) to C(16) and has the highest activity on the C(16) derivative.

GO:0047451 3-hydroxyoctanoyl-[acyl-carrier-protein] dehydratase activity (EC 4.2.1.59)

Comment Specific for 3-hydroxyacyl-[acyl-carrier-protein] derivatives C(6) to C(12).

GO:0047450 crotonoyl-[acyl-carrier-protein] hydratase activity (EC 4.2.1.58)

Comment Specific for short-chain length 3-hydroxyacyl-[acyl-carrier-protein] derivatives C(4) to C(8).

2. The definition of the parent term GO:0019171 3-hydroxyacyl-[acyl-carrier-protein] dehydratase activity should be corrected. The reaction displayed currently is:

CH3-CHOH-CH2-[CH2]n-CO-[acyl-carrier protein] = H2O + CH3-CH-CH-[CH2]n-CO-[acyl-carrier protein].

As displayed about, the hydroxy group is not in the third position as the term name implies. Thus the definition should be changed to the following:

CH3-[CH2]n-CHOH-CH2-CO-[acyl-carrier protein] = H2O + CH3-[CH2]n-CH-CH-CO-[acyl-carrier protein], where n can equal 0.

3. The two GO terms, GO:0080063 3-hydroxybutyryl-[acyl-carrier-protein] dehydratase activity and GO:0047450 crotonoyl-[acyl-carrier-protein] hydratase activity (4.2.1.58) refer to the same activity. These two terms should be merged and should be a child of GO:0019171 : 3-hydroxyacyl-[acyl-carrier-protein] dehydratase activity. The compounds mentioned are synonyms of each other and one of the synonyms (ie. 3-hydroxybutyryl acyl carrier protein dehydratase activity) for the GO term, GO:0047450 is the actual GO term for GO:0080063 (3-hydroxybutyryl-[acyl-carrier-protein] dehydratase activity).

4. The GO terms that are children of GO:4132 fatty acid synthase activity should be changed from having an “is a” relationship to “Part of” relationship, because the fatty acid synthase activity is made up of many individual activities and each of its children alone do not have the fatty acid synthase activity.

Cindy

Reported by: cjkrieger

Original Ticket: geneontology/ontology-requests/7444

[gocentral]

I would also make the GO term GO:0019171 3-hydroxyacyl-[acyl-carrier-protein] dehydratase activity, a child of GO:4132 fatty acid synthase activity, using a "Part of" relationship.

Original comment by: cjkrieger

[gocentral]

I was just reviewing item no. 3 from above and I wanted to suggest that once the two terms, GO:0080063 and GO:0047450 are merged that the name of the GO term should be crotonoyl-[acyl-carrier-protein] hydratase activity because that is the name of the corresponding EC number (EC 4.2.1.58). I also think that the following comment from the EC no. should be added to the definition for the new merged GO term.

Comment: Is specific for short chain-length 3-hydroxyacyl-[acyl-carrier protein] derivatives (C4 to C8).

As mentioned above, the GO term should be made a child of GO:0019171 : 3-hydroxyacyl-[acyl-carrier-protein] dehydratase activity.

Original comment by: cjkrieger

[gocentral]

HI Cindy,

I talked to David about your suggestion #4 yesterday. He thinks that changing the relationships between the terms for the individual activities and the "fatty acid synthase activity" term is fine, provided you feel that it is true that every time the child activity occurs, it is part_of "fatty acid synthase activity". However, before we can do that, we'll need to make sure that every term has an is_a relationship to some other term. Most of them already do, e.g.

[acyl-carrier-protein] S-acetyltransferase activity (GO:0004313) IS_A S-acetyltransferase activity (GO:0016418)

but I think they're a few that do not have an additional is_a relationship to a term that represents the general type of catalytic activity it is.

We can look this together when you're back next week if you like.

-Karen

Original comment by: krchristie

[gocentral]

• assigned_to: nobody --> kchris

Original comment by: krchristie

[gocentral]

I have been doing a lot of enzyme work and have dealt with some of these points already.

2 - GO:0019171 def corrected to use that of the MetaCyc reaction 3 - GO:0080063 merged into GO:0047450; parentage corrected as suggested

1 - MetaCyc has separate entries for each of the reactions that EC:4.2.1.59 and EC:4.2.1.61 catalyze, so I suggest we add those (some of them are already in GO).

4 - I don't think this is a good idea because it confuses the name of a complex ('fatty acid aynthase') with the various reactions that the complex performs. In humans, fatty acid synthase catalyzes a whole host of reactions (see http://BioCyc.org/META/NEW-IMAGE?type=ENZYME&object=CPLX-7628), but in S. cerevisiae, just one (http://BioCyc.org/META/NEW-IMAGE?type=ENZYME&object=CPLX-7627), although the alpha and beta subunits both catalyze a number of reactions. If a GO term really were used to represent what fatty acid synthase is doing, it would need to be a process term like 'fatty acid biosynthesis'.

Original comment by: girlwithglasses

[gocentral]

Hi,

Regarding #4, Peter D'Eustachio is much more knowledgable than I on this, so I will copy in part of an email thread on this subject, below.

I'd also like to add that from an annotator's perspective, it is really a bad idea to put mammalian annotators in a position that when they come across "fatty acid synthase activity" as a catalytic activity, that they will have to know all of the various parts of the fatty acid synthase reaction in order to make function annotations. From the perspective of the single gene product mammalian enzyme, "fatty acid synthase activity" is the appropriate description of its catalytic activity, which I think should belong as a descendent of "catalytic activity". Process is much broader than function, and can include other gene products which affect a catalytic function but do not possess it, so I think it is important for mammalian annotators to be able to find an MF term that represents what the gene they are annotating does.

Regarding correct relationships, Cindy and I talked a bit more. We think that the "has_part" relationship would be the correct statement between the term "fatty acid synthase activity" and the various individual catalytic activities that make it up. That way, when you annotate a mammalian "fatty acid synthase activity" to that term, you indicate that it has those parts, but when you annotate a bacterial gene to a more specific catalytic activity, you make no statement about the individual reaction being part of a larger whole in terms of catalytic activity.

-Karen

On Thu, 26 Aug 2010, D'Eustachio, Peter wrote:

> No, the name looks quite specific - it incorporates all seven (?) steps > of the reaction catalyzed by the multifunctional protein encoded in > yeast and human genomes (and present in cells as a component of a larger > complex), which makes it inappropriate for function annotations of > individual steps catalyzed by the monofunctional E coli proteins. > > Peter > > -----Original Message----- > From: Amelia Ireland [mailto:aji@ebi.ac.uk] > Sent: Thursday, August 26, 2010 12:15 PM > To: D'Eustachio, Peter > Cc: 'Karen Christie'; Chris Mungall; ontology-editors@genome.stanford.edu > Subject: Re: [Ontology-editors] term namespace changes > > > On Aug 25, 2010, at 6:08 PM, D'Eustachio, Peter wrote: > >> A great example of the difficulties. The textbook lists approx seven >> reaction steps in the process of fatty acid synthesis. In E coli, >> there's a separate enzyme encoded by a separate gene for each step. >> In humans (and yeast?) a single large protein with approx seven >> different domains, each with a distinct catalytic function, mediates >> the whole process. So in E coli it's hard not to annotate seven
>> different functions; in humans it's easy to annotate one large one. > > I think that the problem here is that people are looking at the term > name and thinking of what the gene product or complex named 'fatty > acid synthase' does, rather than looking at the definition and seeing > what the term is supposed to represent. Unfortunately the EC accepted > name is not very helpful here because it's so vague... the systematic > name, "acyl-CoA:malonyl-CoA C-acyltransferase (decarboxylating, > oxoacyl- and enoyl-reducing and thioester-hydrolysing)", may not trip > off the tongue, but it does at least provide a good description of > what the reaction being catalyzed is.

Original comment by: krchristie

[gocentral]

Regarding # 2. The def of GO:0019171 in MetaCyc is the same one that GO has/had. The definition needs to be modified so that the hydroxyl (-OH) group is in the beta position to the carboxyl group. I just emailed MetaCyc about the correction and they suggested to make the reaction even more generic by adding an -R to the end of the compound.

The new definition should be: R-[CH2]-CHOH-CH2-CO-[acyl-carrier protein] = H2O + R-[CH2]-CH-CH-CO-[acyl-carrier protein], where n can equal 0.

It sounds like MetaCyc is going to wait and incorporate the changes from GO.

thanks, Cindy

Original comment by: cjkrieger

[gocentral]

Regarding #2, I changed the def to this:

a (3R)-3-hydroxyacyl-[acyl-carrier protein] = H2O + a trans-delta2-enoyl-acyl-[acyl-carrier protein]

I'm not sure whether this change had been made before or after your correspondence with MetaCyc. Does that look OK?

Original comment by: girlwithglasses

[gocentral]

Regarding #2.

You're change to number two looks good, but I think the product would be a trans-delta2-enoyl-[acyl-carrier-protein] and not a trans-delta-enoyl-acyl-[acyl-carrier-protein]. An enoyl means there is a double bond and an acyl so you don't need the extra acyl. I do not know whether all of the products are trans-delta2...We could just say that the products are 2-enoyl-[acyl-carrier-protein]. Just a thought.

Cindy

Original comment by: cjkrieger

[gocentral]

See also SF:3096654, merge fatty-acyl-CoA synthase activity (is a process)

https://sourceforge.net/tracker/?func=detail&aid=3096654&group\_id=36855&atid=440764

Original comment by: girlwithglasses

[gocentral]

Hi Amelia,

Does anything else need doing for this entry, or can we shut it down?

Thanks J,B,P

Original comment by: rebeccafoulger

[gocentral]

• assigned_to: kchris --> girlwithglasses

Original comment by: rebeccafoulger

[gocentral]

• status: open --> closed-out-of-date

Original comment by: rebeccafoulger