Cell Cycle Content Meeting follow-up 12 April 2013
- Jane Lomax
- Val Wood
- David Hill
- Tanya Berardini
- Paola Roncaglia
- Rebecca Foulger
- Prudence M
- Susan Tweedie
- Antonia Lock
- Harold Drabkin
- Chris Mungall
- Mary Dolan
Q: Do we say x process occurs_during y phase?
- Val: Phases are defined based on what gross events people could observe down the microscope. If you add 'occurs_during' it will cause problems for Prokaryotes and some yeast.
- AI: Keep the phase terms in the ontology. Then can use them in XPs.
When Val presents to the GOC tomorrow (Sat 13th April) a keypoint is: remove all annotations to phases as they're not processes. We had alot of annotations to G1 and G2, but they're talking about the transitions, not the phases themselves.
CELL CYCLE TRANSITIONS
There are 4 major cell cycle transitions. GO used to have much more, so we've cleaned these up.
Q: Is the positive feedback loop an integral part of G1/S cell cycle transition, or does it regulate the transition?
- David: Recalls from the meeting that the feedback loop IS integral to the transition, according to Rob.
- Val: Thinks the feedback loop REGULATES G1/S transition.
- GFs regulate CDK4 which signals downstream to regulate transcription. Some gene products that are transcribed FEEDBACK to regulate the transition.
AI: Val to Ask Rob if the feedback loop is part of the transition or if it regulates the transition, to resolve the discrepancy.
NB: G2/M is different mechanism to G1/S because it doesn't include a transcription step. It's all about degradation (see Rob's slides).
SUMMARY: Checkpoints divided into:
- detection phase (e.g. detect damage)
- signal transduction part
We need to add relationships into the ontology. To do this, need to decide: is it the signaling that regulates the response, or the whole checkpoint?
- David: The CHECKPOINT positively regulates the response.
- CHECKPOINTS negatively regulate the phase TRANSITION
- CHECKPOINTS positively regulate the RESPONSE (response = e.g. apoptosis, DNA repair)
- AI: Jane to add in the above regulates relationships. This will deal with many of the outstanding SF items.
We got rid of some checkpoint terms because they didn't make sense, so they were collapsed up to the parent. BUT annotators have therefore lost that granularity. Therefore Q to Chris: Can we get an archived GAF to these specific terms so annotators can re-annotate.
AI: Val will send Chris a list of terms we want archived annotations for (DONE)
MGI have several proteins that are involved in the response (e.g. apoptotic genes including caspase 3, bcl2 etc.) but are wrongly annotated to the checkpoint. tut tut.
David suggest therefore that BEFORE we do any re-annotations, we should rearrange the graph, get the current GPs and see which are annotated to the checkpoint and which are annotated to the response and send them to the experts to see if they think we've got it right.
G0 is controversal: the cell never divides again. Need to look at it again.
15 GO terms currently