annealing -> processes vs. functions

[gocentral]

Hi there,

My first pass for this request, for a new GO:'annealing activity' class as parent of the DNA, RNA, and DNA/RNA strand annealing activity classes, seemed simple enough. As with my other requests, this was especially motivated for text annotation, as it's often difficult to figure out what kind of annealing is being referred to by an isolated mention of annealing.

Then I noticed that GO:'DNA strand annealing activity', defined as "[f]acilitates the base-pairing of complementary single-stranded DNA", has been made obsolete with the comment that "it describes a process and not an activity". However, GO:'RNA strand annealing activity', defined as "[f]acilitates the base-pairing of complementary single-stranded RNA", and GO:'DNA/RNA hybrid annealing activity', defined as "[f]acilitates the base-pairing of single-stranded RNA to double-stranded DNA resulting in the formation of R-loops", are not obsolete. I have to confess that classifying the first as a process and the latter two as functions makes little sense to me given their directly analogous definitions.

I think this points to a larger long-standing issue, that of the nature of the MFs and their relation to the BPs. There are now a number of MF-to-BP part_of assertions, e.g., 'signal transducer activity' part_of 'signal transduction', 'transporter activity' part_of transport. Given these assertions, it seems that MFs are processes since a part of a process would have to be a process itself. Additionally, the (natural-language) definitions of many if not most MF classes are still framed as processes.

It seems that the MF part of GO annotation of genes/gene products intends to specify molecular functionality as realizable entities, since if MFs were just processes, the role of the annotated gene product in the process would be lost; for example, for an entity annotated with transporter activity defined as a process, it could be seen as unclear whether the entity is playing the role of a transporting entity or a transported entity. If MFs are realizable entities, it seems to me a more proper assertional template would be BP realizes MF, e.g., transport realizes 'transporter activity', and I think it'd help to define MF classes along the lines of "the [functional] potential to ..." rather than framing them as processes. However, I think the aforementioned dichotomy among the annealing activity classes would still be an issue.

Anyhoo, this is my two cents. A general annealing activity class would be helpful, though I don't know if it can be accommodated in the current modeling. I realize I've expanded this to some very general ontological issues, so I'm not expecting this to be resolved in the very near future.

Thanks! Mike

Reported by: mbada

Original Ticket: geneontology/ontology-requests/10925

[gocentral]

• labels: --> Other term-related request, New term request
• assigned_to: Paola Roncaglia

Original comment by: paolaroncaglia

[gocentral]

Hi Mike,

Thanks for your comments.

As for the specific 'annealing' issue: The stanza for the obsolete term 'DNA strand annealing activity' contains some additional pointers, not sure if you noticed them:

[Term] id: GO:0000739 name: DNA strand annealing activity namespace: molecular_function def: "OBSOLETE. Facilitates the base-pairing of complementary single-stranded DNA." [GOC:elh] comment: This term was made obsolete because it describes a process and not an activity. is_obsolete: true consider: GO:0000733 [DNA strand renaturation = BP] consider: GO:0036292 [DNA rewinding = BP] consider: GO:0036310 [annealing helicase activity = MF]

Note that annealing helicase activity is part_of DNA strand renaturation

For reference, discussion on (obsoleting) 'DNA strand annealing activity' is recorded here: https://sourceforge.net/p/geneontology/ontology-requests/9608/ In summary, the original 'DNA strand annealing activity' term was okay (or at least better) when it was under NTPase, but once that catalytic information was removed, it just became a process. Hence the obsoletion. In that instance, the other existing annealing terms 'DNA/RNA hybrid annealing activity' and GO:0033592 'RNA strand annealing activity' were been discussed (I checked with Becky). However, I think they're fine as they are. 'DNA/RNA hybrid annealing activity' has one experimental manual annotation and the supporting paper says: "...E. coli Cas3 catalyses ATP-independent annealing of RNA with DNA forming R-loops..." and that looks like a function more than a process. 'RNA strand annealing activity' has 6 experimental manual annotations; the most recent among the supporting papers also leads me to think of function rather than process. (Please see my comments below about the more general 'MF vs BP' issue.)

For the term you need, I added the following:

+id: GO:0097617 +name: annealing activity +namespace: molecular_function +def: "A nucleic acid binding activity that brings together complementary sequences of nucleic acids so that they pair by hydrogen bonds to form a double-stranded polynucleotide." [GOC:mba, http://en.wikipedia.org/wiki/Nucleic_acid_thermodynamics#Annealing] +comment: The term 'annealing' is often used to describe the binding of a DNA probe, or the binding of a primer to a DNA strand during a polymerase chain reaction. The term is also often used to describe the reformation (renaturation) of complementary strands that were separated by heat (thermally denatured). For example, proteins such as RAD52 can help DNA anneal. If at all possible, please annotate to the more specific terms GO:0036310 'annealing helicase activity', GO:0097098 'DNA/RNA hybrid annealing activity' or GO:0033592 'RNA strand annealing activity'. +synonym: "renaturation" RELATED [] +is_a: GO:0003676 ! nucleic acid binding

And I placed 'annealing helicase activity', 'DNA/RNA hybrid annealing activity' and 'RNA strand annealing activity' as is_a children of GO:... annealing activity.

As for the more general MF-BP issue, in brief:

• The traditional view in GO is that MFs are seen as mini-processes, and that's why they're often lexically defined in a way that sounds as such
• We're currently considering an overhaul of the MF branch, and I'll pass your comments on to my other colleagues in this regard. As you say, this won't be resolved in the very near future, so stay tuned :-)

Hope this answers your questions. I'm closing this ticket, but please feel free to add if you have any comments.

Thanks! Paola

Original comment by: paolaroncaglia

[gocentral]

• status: open --> closed-accepted

Original comment by: paolaroncaglia