gene expression

[gocentral]

We've had some discussion about this before, but we think it would be useful to add a new high-level term 'gene expression'. Perhps like this:

biological process ---[i] gene expression ; GO:new ------[p] transcription ------[p] translation ------[p] regulation of gene expression ; GO:new ---------[i] regulation of gene expression, post-translational ; GO:new

any thoughts?

Reported by: jl242

Original Ticket: "geneontology/ontology-requests/3148":https://sourceforge.net/p/geneontology/ontology-requests/3148

[gocentral]

Logged In: YES user_id=631592

I've thought about this a lot. It has lots of issues associated with it, but the HUGE argument for it is if we combine the description of a phenotype with GO Process, many phenotypes are categorized by altered gene expression.

If we do add it, we really need to be all-inclusive. Transcription itself won't suffice because mRNA half life and protein half life also play into this.

David

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=451873

I don't think it would be as problematic to create this term as we might imagine. All the processes that regulate gene expression can go as children of 'regulation of gene expression' e.g.

regulation of gene expression ; GO:new ---[i] regulation of gene expression, post-translational ; GO:new ------[i] regulation of mRNA stability ; GO:0043488

'regulation of transcription' would also be a child of 'regulation of gene expression'...

I think the biggest problem would be ensuring we had all relevant terms as children of 'gene expression' without causing tpvs. But because gene expression is such an all encompassing term, I don't think this would be too much of a problem...

Original comment by: jl242

[gocentral]

Logged In: YES user_id=451873

I don't think it would be as problematic to create this term as we might imagine. All the processes that regulate gene expression can go as children of 'regulation of gene expression' e.g.

regulation of gene expression ; GO:new ---[i] regulation of gene expression, post-translational ; GO:new ------[i] regulation of mRNA stability ; GO:0043488

'regulation of transcription' would also be a child of 'regulation of gene expression'...

I think the biggest problem would be ensuring we had all relevant terms as children of 'gene expression' without causing tpvs. But because gene expression is such an all encompassing term, I don't think this would be too much of a problem...

Original comment by: jl242

[gocentral]

Logged In: YES user_id=561361

Looks fine to me.

-Pankaj

Original comment by: jaiswalp

[gocentral]

Logged In: YES user_id=803460

I like the proposal. Also, "regulation of gene expression, epigenetic" and its "negative/positive" children would have a home!

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=546388

I'm being a devil's advocate here, but I think the gene expression term and it's regulation are as bad as the "biosynthesis of" terms. If you are annotating to these broad terms, it means the experiments are poorly designed and inconclusvie. Especially if everything (transcription, translation, modification, etc.) falls under them. So, if someone treats with something and sees the level of something increase, it is meaningless to annotate to "positive regulation of x biosynthesis or expression". If I were a paper reviewer, I would send the paper back and say it needs another experiment or two to find out what exactly IS happening. It would be better to have no annotation at all.

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=436423

I'm entirely sympathetic to concerns about 'gene expression' or 'regulation of gene expression' being used unwisely in annotations. On the other hand, it seems that we do need the terms to organize the ontology at a fairly high level. Food for thought ...

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=451873

And as David says, this term is useful for things outside of annotation, like cross products with other ontologies. We would just need to be careful it wasn't used for bad annotations.

j

Original comment by: jl242

[gocentral]

Logged In: YES user_id=473890

Hi,

I think having the 'gene expression' and 'regulation of gene expression' terms are reasonable. When I was in graduate school at Berkeley, a number of professors, including my advisor, used these types of terms to describe all of the ways that the expression of a gene might be controlled, so I think these terms are consistent with the way bench scientists speak of these processes.

I understand Harold's issue with using such broad terms for annotation, but I think that's a different issue relating to annotation practice, and shouldn't be a reason not to put these terms into the ontology.

-Karen

Original comment by: krchristie

[gocentral]

Logged In: YES user_id=803460 Originator: NO

Can we make this higher priority? We need this term to annotate some phenotypes with PATO.

Thanks, Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=254723 Originator: NO

We need it.

What would the definition be? I'm worried by it being "all-encompassing" - the term should only mean one thing (it can still be a very general definition of course). As David often points out, we have to think about timing - over what temporal duration does gene expression unfold? Different organisms will have different process parts but that's ok.

Would a single instance of a gene expression process/event create exactly one RNA?

We should look at the definition of terms in SO here and check to see that if we were to import SO's definitions wholesale, would they work for us?

Original comment by: cmungall

[gocentral]

Logged In: YES user_id=803460 Originator: NO

wikipedia had "Gene expression, or simply expression, is the process by which a gene's DNA sequence is converted into the structures and functions of a cell. Non-protein coding genes (e.g. rRNA genes, tRNA genes) are not translated into protein."

I think we could have something along those lines: "The processes by which a gene's coding sequence is converted into a functional gene product (protein or RNA). This includes the production of messenger RNA for protein-coding genes."

Also, I think this term should be a parent of GO:0006350 : transcription and GO:0043043 : peptide biosynthesis.

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=473890 Originator: NO

The wikipedia def Pascale mentioned is consistent with the way I've heard people use this term. My take on it was that by "gene expression" people meant the entire process by which a gene product was produced, or expressed, from the corresponding gene, definitely including translation to produce the actual protein/peptide. In the usage I encountered, "regulation of gene expression" definitely included processes that regulated the translation of an mRNA to regulate how much protein was actually produced in the cell.

-Karen

Original comment by: krchristie

[gocentral]

Logged In: YES user_id=451873 Originator: YES

I like Pascale's definition.

One transcriptional event may generate more than one gp though, because the mRNA may be spliced, so we could ammend the def as follows:

"The processes by which a gene's coding sequence is converted into a functional gene product or products (proteins or RNA). This includes the production of messenger RNA (mRNA) for protein-coding genes, and the translation of that mRNA into protein."

(also added a bit about translation).

Q: should gene expression include post-translational modifications to proteins or not?

Jane

Original comment by: jl242

[gocentral]

• assigned_to: nobody --> jl242

Original comment by: jl242

[gocentral]

Logged In: YES user_id=631592 Originator: NO

Q: should gene expression include post-translational modifications to proteins or not?

I think by definition they have to be if they are required for the GP to function. For example, some proteases and neuropeptides are translated as precursors and can't function until they are processed. Trickier are cases where a protein kinase is active only in its phosphorylated form.

How about:

"The processes by which a gene's coding sequence is converted into a mature gene product or products (proteins or RNA). This includes the production of messenger RNA (mRNA) for protein-coding genes, and the translation of that mRNA into protein."

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=451873 Originator: YES

Yep, I like that definition.

Original comment by: jl242

[gocentral]

• milestone: --> 224410

Original comment by: jl242

[gocentral]

Logged In: YES user_id=546388 Originator: NO

Although I can understand Midori's point about the need for an organizing term (grouping), I would be still concerned that this term would be applied to any situation in which the "amount" of something has changed due to an experiment. Some gene product is involved in gene expression because if you mutate it or knock it out, the level/amount/expression of the gene product of some other gene changes. End of experiment. Can this be useful at all? I came across a term the other day that was being used in an proposed annotation in a test set: GO:0065009, regulation of a molecular function. Don't know what function, but the protein was annotated to it.

I assume that the term GO:0065009 was established to provide an is_a path. But it shouldn't be used to annotate to. Protein x is involved in gene expression. Expression of what? Perhaps it might be useful if there was a requirement that if this term is used, the WITH field must be filled in with the target.

Original comment by: hdrabkin

[gocentral]

• milestone: 224410 -->

Original comment by: hdrabkin

[gocentral]

Logged In: YES user_id=473890 Originator: NO

I mostly like Pascale's proposed def, but the second sentence is protein-coding gene centric. Maybe we could broaden it a bit.

"The processes by which a gene's coding sequence is converted into a mature gene product or products (proteins or RNA). This includes the production of an RNA transcript as well as any processing to produce a mature RNA product or an mRNA (for protein-coding genes) and the translation of that mRNA into protein. Some protein processing events may be included when they are required to form an active form of a product from an inactive precursor form."

While I understand Harold's concern, I still think this is an annotation issue that shouldn't prevent adding a term to the ontology if it improves our representation of the biology.

I'd also like to add that I think his proposal to require the WITH field to be filled in with a target is a very different use of the with field than what we have allowed here before. This would NOT be supporting evidence in the same way we have used this field for other uses, where the thing in the with field is something that the annotated gene is being compared to in order to make the annotation. I am against this usage for the with field.

-Karen

Original comment by: krchristie

[gocentral]

Logged In: YES user_id=436423 Originator: NO

I agree with Karen about the 'with' field. I could happily live with either of the latest two proposed defs.

m

Original comment by: mah11

[gocentral]

Logged In: YES user_id=1320989 Originator: NO

Hi,

do you know what's happening with this item? Is it stalled? I'd like to use the regulation of gene expression term.

Chris

Original comment by: c_tissier

[gocentral]

Logged In: YES user_id=803460 Originator: NO

I think this should be implemented.

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=463625 Originator: NO

Hi jane et al,

heres another paper i came across today where this term would be useful... PMID:10649442

cheers Ev

Original comment by: ecamon

[gocentral]

Logged In: YES user_id=631592 Originator: NO

Hi Everyone,

As you may know, Tanya and I are working on resolving issues associated with the regulation terms in the biological process ontology. We have come to the term 'regulation of gene expression, epigenetic' that needs a home, so we think it is time to implement this SF item. Here is what we propose. We would like to implement this almost exactly as is outlined here with the following slight change:

biological process ---[i] gene expression ; GO:new ------[p] transcription ------[p] translation ------[p] protein maturation ; GO:0051604 ------[p] RNA processing ; GO:0006396 ------[p] regulation of gene expression ; GO:new ---------[i] regulation of gene expression, epigenetic ; GO:0040029 ---------[i] circadian regulation of gene expression ; GO:0032922

regulation of biological process ------[i] regulation of gene expression ; GO:new

GO:new1, gene expression def: "The process by which a gene's coding sequence is converted into a mature gene product or products (proteins or RNA). This includes the production of an RNA transcript as well as any processing to produce a mature RNA product or an mRNA (for protein-coding genes) and the translation of that mRNA into protein. Some protein processing events may be included when they are required to form an active form of a product from an inactive precursor form."

GO:new2, regulation of gene expression def: "Any process that modulates the frequency, rate or extent of gene expression. Gene expression is the process in which a gene's coding sequence is converted into a mature gene product or products (proteins or RNA). This includes the production of an RNA transcript as well as any processing to produce a mature RNA product or an mRNA (for protein-coding genes) and the translation of that mRNA into protein. Some protein processing events may be included when they are required to form an active form of a product from an inactive precursor form."

We should make a related synonym for 'RNA processing' called 'RNA maturation'.

We think we should work with the annotation group to develop strict guidelines on the use of GOnew1 and GOnew2. We may want to revisit the annotations in 6 mo (after implementation) to see who has used this term and how.

We plan to implement this by November 16. Please let us know if you have comments.

David and Tanya

Original comment by: ukemi

[gocentral]

Logged In: YES user_id=803460 Originator: NO

look good! thanks for dealing with that one.

I suggest you add the terms here: http://gocwiki.geneontology.org/index.php/Misused\_terms and possibly we can link to the GO term wiki (Jim's) and add more specific comments there?

Pascale

Original comment by: pgaudet

[gocentral]

Logged In: YES user_id=579762 Originator: NO

Added today:

gene expression, GO:0010467 regulation of gene expression, GO:0010468

revision 5.554

Added terms to list of (potentially) misused terms.

When the terms get absorbed into Jim's wiki, we will add comments there.

Original comment by: tberardini

[gocentral]

• status: open --> closed-accepted

Original comment by: tberardini