Transitive over part of documentation

From GO Wiki
Jump to navigation Jump to search

File to refer to:

[mp-asserted.doc]


Email thread:

Chris,

Tanya and I have started to go through these here is one that seems funny:

GO:0000706 "meiotic DNA double-strand break processing" PART_OF GO:0006311 "meiotic gene conversion" GO:0000706 "meiotic DNA double-strand break processing" PART_OF GO:0007131 "meiotic recombination"

The issue here GO:0000706 is both a part of GO:0006311 and GO:0007131. But GO:0006311 and GO:0007131 are independent processes, so every time there is an instance of GO:0000706 it is EITHER GO:0006311 or GO:0007131 but not both. I think this points out an error in the ontology as opposed to an error in the transitive over part_of rule. Can you generate a separate list of all the terms that have more than 1 part_of relationship?

T & D


Chris and Tanya,

Actually, from looking at the structure of the graph, it looks like GO:0006311 "meiotic gene conversion" could actually be an is_a GO:0007131 "meiotic recombination". What do you think? I'm trying to see what I can glean from a textbook. Maybe we should check with a recombination person, maybe Eurie. Seems like there are a lot of these in this part of the graph. In any case, it seems like there is an issue here that needs to be fixed.


David


Hmmmm, I don't think so.

Look at the stanzas and the exact synonyms in particular:

[Term]
id: GO:0006311
name: meiotic gene conversion
namespace: biological_process
def: "The cell cycle process whereby genetic information is transferred from one helix to another. It often occurs in association with general genetic recombination events, and is believed to be a straightforward consequence of the mechanisms of general recombination and DNA repair. For example, meiosis might yield three copies of the maternal version of an allele and only one copy of the paternal allele, indicating that one of the two copies of the paternal allele has been changed to a copy of the maternal allele." [ISBN:0815316194]
synonym: "gene conversion without reciprocal crossover" EXACT []
is_a: GO:0006310 ! DNA recombination
is_a: GO:0022402 ! cell cycle process
relationship: part_of GO:0007127 ! meiosis I

[Term]
id: GO:0007131
name: meiotic recombination
namespace: biological_process
alt_id: GO:0000021
alt_id: GO:0007145
def: "The cell cycle process whereby double strand breaks are formed and repaired through a double Holliday junction intermediate. This results in the equal exchange of genetic material between non-sister chromatids in a pair of homologous chromosomes. These reciprocal recombinant products ensure the proper segregation of homologous chromosomes during meiosis I and create genetic diversity." [PMID:2087779]
synonym: "female meiotic recombination" NARROW []
synonym: "gene conversion with reciprocal crossover" EXACT []
is_a: GO:0006310 ! DNA recombination
is_a: GO:0022402 ! cell cycle process
relationship: part_of GO:0007127 ! meiosis I


So:

meiotic gene conversion = gene conversion without reciprocal crossover
meiotic recombination = gene conversion with reciprocal crossover

I think we need two new terms:

'meiotic DNA double-strand break processing involved in meiotic gene conversion'

and

'meiotic DNA double-strand break processing involved in meiotic recombination'

> I'm trying to > see what I can glean from a textbook. Maybe we should check with a > recombination person, maybe Eurie. Seems like there are a lot of these > in this part of the graph. > In any case, it seems like there is an issue here that needs to be fixed.

Yup. Looking at the list of all terms like this should help us to clean up the ontology some more.

Tanya


seems like we need a way to say: these two processes have no parts in common, to help spot these.

(this is separate from a normal disjoint_from, which says, perhaps confusingly, that nothing can instantiate both these terms)

Chris


Hi Tanya and Chris,

Yes, I think you a right, but I think the term 'meiotic recombination' is a bad name. Maybe it should be 'meiotic recombination with reciprocal crossover'. I agree that the best way to disambiguate this would be to create

'meiotic DNA double-strand break processing involved in meiotic gene conversion'

and

'meiotic DNA double-strand break processing involved in meiotic recombination'

as you suggest, but it seems that we would need a parent for these that would be 'meiotic DNA double-strand break processing'. This term should then relate to the two new terms via is_a relationships and I think is should have a part_of relationship to some term that describes all types of recombination that can take place during meiosis, meiotic recombination???? I just looked at the list that Chris sent and it seems like some of the component ones are legit, but the process ones are remnants of 'has_part' relationships in the ontology. When we first started we didn't distinguish is_part from has_part.

So, I think we should put this on our to-do list before we create the transitive over part_of regulation links. Man, our list just keeps getting longer, but it's good.

David


Comment (Feb. 13 '08): Val independently noticed the problem with these terms and submitted a SF item, which is assigned to Eurie:

https://sourceforge.net/tracker/index.php?func=detail&aid=1059429&group_id=36855&atid=440764

-midori