Ontology meeting 2015-02-26
Attendees: David H, Harold, Heiko, Judy, Paola, David OS, Huaiyu, Paul T
Minutes: Paola
Issues With Ion Channels
Right now the ion-specific channels are not defined as being selective for a type of ion. Therefore, non-selective channels are being annotated to the ion specified terms. In the long run this is misleading and is not intuitive to electrophysiologists.
[Paola's thoughts] Currently we don’t distinguish between selective and non-selective ion channel, in channel activity terms:
GO:0005216 ion channel activity = Catalysis of facilitated diffusion of an ion (by an energy-independent process) by passage through a transmembrane aqueous pore or channel without evidence for a carrier-mediated mechanism.
However, we imply that ion channels are selective when it comes to other terms:
GO:0034702 ion channel complex = A protein complex that spans a membrane and forms a water-filled channel across the phospholipid bilayer allowing selective ion transport down its electrochemical gradient.
GO:0044325 ion channel binding = Interacting selectively and non-covalently with one or more specific sites on an ion channel, a protein complex that spans a membrane and forms a water-filled channel across the phospholipid bilayer allowing selective ion transport down its electrochemical gradient.
Solutions:
1) Annotate a non-selective channel to more than one term to indicate the multiple activities that a given channel may have. E.g. transient receptor potential channels are non-selective cation channels that let many types of cations through, mainly Na+, K+ and Ca2+. Such a TRP channel would be annotated to all three sodium, potassium and calcium channel activities. Drawback: fine if you’re looking at annotations for a single gene product and you have the background knowledge that seeing 3 different channel activities means the channel is non-selective; otherwise confusing. Same if you’re asking an electrophysiologist with limited knowledge of the GO. Also, if you’re doing an enrichment analysis, the results wouldn’t be immediate to interpret.
2) Create children of ion channel terms to indicate selectivity. E.g.
chloride channel activity
---(is_a) selective chloride channel activity
---(is-a) non-selective chloride channel activity
But we already have loads of terms under ion channel activity. To avoid unnecessary combinatorial explosion, we should only create selective and/or non-selective children for the types of channels where this distinction is currently known. Then, annotate to either the selective or non-selective ion channel activity term, and to one of the existing terms that possess granularity of other type as appropriate (e.g. histamine-gated chloride channel activity, mechanically-gated ion channel activity).
How are known non-selective channels currently represented? E.g. transient receptor potential channels are non-selective cation channels that let many types of cations through, mainly Na+, K+ and Ca2+. Currently, TRP and TRPG (fly) are annotated to calcium channel activity. TRPL (fly) to calcium and cation activity terms. Is this satisfactory or are we losing information? I’d say the latter, because for TRP and TRPG I’d read the annotation as ‘gene product lets calcium through’ and if I didn’t have previous knowledge of TRP channels, I’d ignore that they also let other cations through. For TRPL, I’d interpret the dual annotation as the cation one being redundant, rather than implying that other cations can pass.
What selective and non-selective terms would we need? What’s the state of the art knowledge on selective and non-selective terms. (E.g. http://en.wikipedia.org/wiki/Ion_channel#Classification_by_type_of_ions) This list would also enable us to look at existing annotations that would need to be updated.
3) A simplified compromise would be to only add 4 new terms:
selective cation channel activity
non-selective cation channel activity
selective anion channel activity
non-selective anion channel activity
Then annotate to either the selective or non-selective cation/anion channel activity term, and to one of the existing terms that possess granularity of other type as appropriate.
DISCUSSION AND ACTION ITEMS:
- Create non-selective anion/cation channel activity terms (Paola)
- Current terms: do not change the name, but write ‘selective’ in the def, and add e.g. ‘calcium-specific’ as synonym (Paola)
- Check that for all existing cross-products, the genus should be the ‘selective’ term (i.e. the existing one) (Paola). Note: these ontology changes must be in place before the annotation call below.
(Update: all this is in a SF ticket now: https://sourceforge.net/p/geneontology/ontology-requests/11540/)
- Will need to work with annotation group to reannotate. Can paint provide a report? Yes they can look into gene families.
- David H’s plan (approved): we explain this on an annotation call; (update: Rama suggested April) we can suggest to look specifically at TRP channels, but other than that, databases need to look at all their own annotations to ion channels. (Update: Huaiyu composed a list for calcium channels, see his email today.) We will also need to provide some guidance on annotation extensions, and on looking for the ‘right’ ion to annotate to, rather than for ‘unnatural’ ions used in assays (e.g. rubidium). (David H and Huayiu)
- Then paint can serve as a second check, by suggesting findings from their gene families.
What do we want to do about intergenic regions?
All agree that we should *not* have terms for transcription from intergenic regions, as if a function is found for the resulting transcripts, we wouldn't call the region intergenic any more Anyway - the paper this request is based on is not really about intergenic regions, it is about control of production of untranslated transcripts associated with TSS and TTS of protein coding genes. Solution: make term for transcription of non-coding transcripts associated with TSS/TTS of protein coding genes. Can then make term for regulation of this transcription. Need a snappier names for these these though.
How to define protein families in enzyme binding terms
This follows up from a previous call where we resolved: "We have decided that we will include families. How we define families will need to be addresses. Is it only if there is more than one similar gene in the same organism? The bottom line here is that a gene protein is not binding the activity, it is binding the molecule."
We need to address how to define families.
Replacing has_regulation_target with more specific, chain based relations
In discussion with annotators (Ruth, Rachael, Becky, Aleks) - agreed that the confusing annotation extension relations - has_regulation_target - should be replaced by more specific relations defined using property chains. We already have some of these, but the names, which embed the property chain, are confusing to most people.
What relations to have and what to call them?
regulates_o_has_agent -> regulates_activity_of - with restriction that only applies to regulation of MFs (BPs having many agents) ? regulates_o_has_substrate -> ?
How deep to go?
reguates_transport_of ? regulates_transcription_of ? <- This might seem a step too far, but it would be much clearer than the current confusing mess.
results_in strategy
See emails from Chris following last week's call (threads on go-ontology 'results_in outliers' and 'results_jn refactor'). Chris asked for feedback on his proposal. Discuss, then reply to email thread.
BTW, Antonia recently requested this on TG: 'positive regulation of pyrimidine-containing compound salvage by positive regulation of transcription from RNA polymerase II promoter'. Put on hold until we decide...