Closed ValWood closed 8 years ago
mah11
commented
8 years ago Yes, these are experimental techniques rather than conditions.
Strictly speaking, they fit the scope of OBI (Ontology for Biomedical Investigations; http://obofoundry.org/ontology/obi.html) but that ontology is one scary f*%^er. I'd rather leave the info in comments - and I've done similar occasionally - than have to deal directly with OBI in Canto.
ValWood
commented
8 years ago they really do look like shmooing cells! but I agree I kept as tapered, but they look different from other tapered in a subtle way (although haven't we wondered before if the tapered phenotype of the mitochondrial mutants is actually a microtubule defect...?)
added one phenotype FYPO:0001018 abolished activation of bipolar cell growth
deleted penetrance=high from normal phenotypes (these might be copy edit errors, I keep leaving previous extensions in when I copy paste).
positive regulation of protein localization to cell tip I agree with your comment. This should not be “regulation of” s it is directly involved (required). GO is a bit weird, in that the regulation terms are not necessarily more specific. so these are now protein localization to cell tip
[ ] I will try to find some documentation at GO to explain this.
changed the allele name and description of tip1Δ299 from name(description) tip1(Δ299-461) tip1Δ299(299-461) to follow standard for partial amino acid deletion description
Does that make sense?
[x] Q should human ortholog be? DCTN1
[x] Add a "misc" annotation in Artemis tip1 localisation to cell end dependent on microtubule cytoskeleto
changed my legacy GO annotation from GO:0071963 establishment or maintenance of cell polarity regulating cell shape to GO:0071964 establishment of cell polarity regulating cell shape does that sound correct?
Nice session, and paper! I remember reading that one years ago to make my 2 GO annotations!
ValWood
commented
8 years ago /curs/b19dfb2b27f82b53 now through when new terms added. No comments on this session
ValWood
commented
8 years ago pombe/curs/f2b0ffda72370b22
I am replacing my old GO annotations tea1 GO:0034613 cellular protein localization IMP has_input PomBase:tip1 by more specific tea1 GO:1990896 protein localization to cell cortex of cell tip IMP has_input pom1
This will often apply with the legacy GO annotation. Usually I make a new annotation, to see if I can "go deeper" Then remove the old annotation. So if you need to, you can ask me to delete the old annotation, or I will spot them and delete the old one it when we approve.
These sessions are spot on so I will stop checking them properly soon :) I'll just do a cursory check for obvious stuff like copy paste extensions etc.
ValWood
commented
8 years ago https://curation.pombase.org/pombe/curs/f2b0ffda72370b22#
Also spot on. I only changed my GO annotations.
I also removed the GFP and YFP from the geneotypes.
When we capture these, if the GFP is presumed not to affect the protein we only capture the allele of the mutated gene. We don't bother with the GFP of the WT localized gene.
If we did, these would all be multi-gene phenotypes, rather than single gene phenotypes.
ValWood
commented
8 years ago curs/ce1fc9c90e999884 PMID:15177031 also spot on.
For future sessions do you want to put the phenotype term request straight into the FYPO tracker. https://github.com/pombase/fypo/issues
This way I will start approve the sessions once the FYPO terms have been created, then I can add the new terms while I am checking, which will save time.
ValWood
commented
8 years ago I can't find any documentation on the GO site about "regulation" relationship and terms.
Basically, in the ontology we always try to annotate as specifically as possible. Most terms are related by is_a and Part_of but all regulation terms are related by the relation "regulates" As illustrated here: http://www.ebi.ac.uk/QuickGO/GTerm?id=GO:0010695#term=ancchart
So we always annotate as specific as possible, but if we don't know if a gene product is "regulating", or "directly involved" we stick with the parent.
A gene product might be both directly involved and regulating (for example a rate limiting step in a metabolic pathway) Or neither (just a phenotype because an upstream process is compromised).
It took us a long, long time to figure out when to use the term itself and when to use regulates. We still don't have good guidelines but we usually know. We sometimes check with each other if we aren't sure.....
It would be good to have your thoughts on this at a future curation meeting. There are still some of the cell cycle signalling where I am not sure.....
So I'm going to close this ticket, as any tasks are done and the rest was just for reference.
I'll document changes I made here, in case anyone can correct me, or we can make improvement on what I have done:
I'm moving all of these to comments. To me, these sound like method details. If anywhere they would be subtypes of ECO terms perhaps? I'll keep the note in comment in case they are useful later if we have more specific evidence codes.