Cell Cycle Content Meeting follow-up 27 March 2013: Difference between revisions

From GO Wiki
Jump to navigation Jump to search
Line 15: Line 15:
  ** [i] cell cycle phase
  ** [i] cell cycle phase
  *** [i] developmental phase, etc
  *** [i] developmental phase, etc




Line 22: Line 21:
  * Propose we add a soft check and a comment, "not to be used for direct annotation", and refer to the existing part of children for reannotation
  * Propose we add a soft check and a comment, "not to be used for direct annotation", and refer to the existing part of children for reannotation
  * Defs currently make these sound like processes, redefine as phases?
  * Defs currently make these sound like processes, redefine as phases?
   
  *N Val has sent lists of phase annotations to the  DBs present at the meeting : Pombase (updated) , Uniport, flybase, TAIR (updated), MGI  plus SGD
 
4. Are we going to obsolete the "regulation of phase terms" (cleaner)
 
Votes for obsoletion Val; Chris
 
All:
As you reannotate these could you provide the replacement terms to be added as comments to the obsoleted terms
e.g, regulation of
S-phase is usually regulation of DNA replication
 
 
5.  Val Editors need to delete the part_of  links between the phase terms and other children
 
Chris Do you mean replace the part_of links between processes 9e.g. transitions) and phases to be 'during' links? I had imagined keeping them as part_of at first.
I'd really like to see logical definitions for the transition processes in terms of the phases. This might need a richer set of relations. E.g
GO:0000216 ! M/G1 transition of mitotic cell cycle
EquivalentTo
'mitotic cell cycle phase transition' and starts_during 'M phase' and ends_during 'G1 phase'
 
We need to discuss this further.... (val I guess it would be OK if people were prevented from making direct annotations to the phase terms as a start, but there are other problems (more detail requited here....,) I think part_of isn't the right relationship (and causes a lot of TPVs)
 
 
 
 
6. Reword the phase definitions so that it would not make sense logically for the regulation of transitions to be part_of children of phases.
 
==== mitosis and meiosis I & II ====
==== mitosis and meiosis I & II ====



Revision as of 11:59, 26 March 2013

Meeting report

  • Work through comments

Phase terms, TO DO

phase terms relationships to cell cycle events etc

1. The relationship between the phases and other cell cycle processes. David OS has suggested that we could continue to use part_of, or when part_of is too strong use happens_during (also starts_during and ends_during). But I got the sense during the meeting that we wanted to steer away from using part_of between processes and phases and make the phases both is_a and part_of disjoint from the rest of bp? The main part_of children are currently the cell cycle transitions.

 * Val: I feel *very strongly* that they should all be happens_during....I think we should stick with this  *unless* anyone can find an annotation where this is not true... 

2. How to handle phases that are also processes. It's possible to frame some phases as processes, indeed some phases are defined this way e.g. metaphase. Would we want these to be represented under both 'cell cycle phase' and 'cell cycle process' and have two separate terms? Or do we decide which should be phases and which should be processes and just split them up?

3. Do we want to make this node generic for all phases in GO, and use it to house some of the e.g. developmental phases there? How might that look e.g.:

* biological process
** [i] cell cycle phase
*** [i] developmental phase, etc


* I have asked Chris for a list of direct annotations
   * Chris has sent this list and there are only 243 for all terms for all 45 terms for all databases after the inferred annotations are removed... I have sent lists to PomBase,MGI,Uniprot,flybase,TAIR,SGD which will leave 43.... 
* Propose we add a soft check and a comment, "not to be used for direct annotation", and refer to the existing part of children for reannotation
* Defs currently make these sound like processes, redefine as phases?
*N Val has sent lists of phase annotations to the  DBs present at the meeting : Pombase (updated) , Uniport, flybase, TAIR (updated), MGI  plus SGD

4. Are we going to obsolete the "regulation of phase terms" (cleaner)

Votes for obsoletion Val; Chris

All: As you reannotate these could you provide the replacement terms to be added as comments to the obsoleted terms e.g, regulation of S-phase is usually regulation of DNA replication


5. Val Editors need to delete the part_of links between the phase terms and other children

Chris Do you mean replace the part_of links between processes 9e.g. transitions) and phases to be 'during' links? I had imagined keeping them as part_of at first. I'd really like to see logical definitions for the transition processes in terms of the phases. This might need a richer set of relations. E.g GO:0000216 ! M/G1 transition of mitotic cell cycle EquivalentTo 'mitotic cell cycle phase transition' and starts_during 'M phase' and ends_during 'G1 phase'

We need to discuss this further.... (val I guess it would be OK if people were prevented from making direct annotations to the phase terms as a start, but there are other problems (more detail requited here....,) I think part_of isn't the right relationship (and causes a lot of TPVs)



6. Reword the phase definitions so that it would not make sense logically for the regulation of transitions to be part_of children of phases.

mitosis and meiosis I & II

  • outstanding question: although this is currently a little odd, because although mitosis has a nuclear division parent, meiosis meiosis I and meiosis II don't....maybe the meiosis terms are a little broader and encompass MORE than nuclear division ? although maybe this is confusion between a "meiotic cell cycle" and "meiosis".... I'm not sure.....anyone else know?
  • follow up to Q above: OK meiosis is "a specialised type of nuclear division" so it should also have this parent (I'll ask Midori if there is a reason why it doesn't). When people refer to "meiosis" in the broader sense, they are really talking about the meiotic cell cycle, or meiotic development (i.e which would include sporulation in fission yeast). We should probably define meiosis and mitosis as a type of nuclear division (qualified by the cell cycle in which it takes place (meiotic or mitotic). each should have 2 parents the "x cell cycle" & "nuclear division". I'm struggling with meiosis I and meiosis II, because if these are the specialised nuclear divisions there are lots of child processes which are not part of nuclear division (DNA replication processes and the G2/M transition for example). It seems as though we need a term which represent to represent all of the events which occur (meiosis I + interphase) ......I have not seen such phrases used ........and I haven't managed to find anything yet....?

(Midori thinks the ommission was just an ommission)

  • more follow up: OK it sounds as though we need to add these "nuclear division" parents, and then find a way to represent "meiosis I cell cycle process" and "meiosis II cell cycle process" (maybe that is the way?) to capture all the events which happen during interphase AND meiosis of those specific divisions, and then move the existing children of meiosis I & II to these new parent terms where appropriate (obviously the events which are analogous to the mitotic events can stay under the Meiosis I and meiosis II terms). Would that work?

Transitions TO DO

Question from Val:

I think we might have a problem with the G1/S transition def (for pombe, it seems a little specific. and it doesn't really match the G2/M definition if it is only describing the switch itself should be more like:

The mitotic cell cycle transition by which a cell in G1 commits to S phase. The process begins when the kinase activity of G1 cyclin-dependent kinase (G1 CDK), reaches a threshold high enough for the cell cycle to proceed. This is accomplished by activating a positive feedback loop that results in the accumulation of unphosphorylated and active G1 cyclin-dependent kinase (G1 CDK).

The way I understood the discussion 1. The CDK cyclin complex activation was the transition (i.e a switch) so this process is actually a single molecular event ( activation of the CDK, and the only things which should be annotated to the transition itself were CDK and the cyclin 2. There are numerous ways the switches can be positively and negatively regulated

G1/S i) CDK itself positively regulates G/M transcription of Cyclins (via MBF) as part of +ve feedback loop

ii) CDK negatively regulates transition by inhibiiton of SIC1 G2/M

i) CDK itself positively regulates G/M transcription of Cyclins as part of +ve feedback loop (NOT In pombe)

ii) cdc25 (and any upstream activators of) positively regulates transition via dephosphorylation

iii) wee1 and any upstream activators of negatively regulate via phosphorylation


Can anyone confirm this or correct me? If no one is sure, I will check with the experts (vw: not yet done)

Checkpoints TO DO

  • G2 DNA damage checkpoint: this is should only be a grouping term for mitotic and meiotic G2 DNA damage, which should always be annotatable, so :

i) should flag as not to be used in direct annotation ii) add missing children like signal transduction involved in DNA damages checkpoint

  • Merges of the checkpoint terms. To demerge or not.
  • Regulation between checkpoint signaling and checkpoint response (signaling is_a positive regulation of response?)
  • Regulation between checkpoint signaling and phase transition (signaling is_a negative regulation of phase transition?)
  • Regulation between checkpoint and phase transition (checkpoint is_a negative regulation of phase transition?)

negative regulation of Checkpoints

The only *known* negative regulation so far is DNA damage checkpoint https://sourceforge.net/tracker/?func=detail&aid=3608141&group_id=36855&atid=440764 a suggestion, should we obsolete the other terms, they could be re-instated if this type of regulation is found to exist?

Response from Jane : I actually just added these regulation terms so I could merge 'regulation of G1/S transition checkpoint' etc into them (if you remember we decided to do away with the checkpoints classified only by the phase).So I don't have specific examples - we could add this to the obsolete list but perhaps we want to run it by the experts? Are the other regulation of checkpoints terms valid (i.e. regulation and positive regulation?)

Further question from Val : We have only decribed detection, and the signalling pathway. I am not sure how "positive regulation " would fit in here....I'm not sure that this is needed (unless it is detection).....D/T what do you think?

SF tickets

  • regulation of CDK activity [1]