Closed CuzickA closed 5 years ago
what else fits in here?
Sorry, I don't know this biology well enough to help with this question.
although aiming for species neutral phenotypes we will also need some species specific phenotypes such as filamentous growth -does this seem ok?
Yes, as long as a species-neutral phenotype is never a subclass or part of a species-specific phenotype.
should reproductive phenotypes be added here or kept separate?
I'm inclined to keep them separate (as they are in GO).
Any abnormal biological process, morphology or function not related to pathogenicity will probably go under here.
You will also probably have a lot of plant structure related terms (these not covered by GO). I presume there PO can be used to define these , but maybe leave those for a little while and concentrate on the most used terms and especially the pathogen-host interactions.
[ ] I would make your stature phenotype into a plant-specific term, and place under a grouping term for plant anatomy term. Would "plant stature" be a "plant anatomical phenotype"?
[ ] How is the "yeast-like growth form phenotype" described? Growth is horrendously tricky, (for me...it still confuses me after 20 years). Are you referring to i) an increase in cell size (in this case I think this can be a species agnostic "cell growth" term) or ii) a yeast specific growth "pattern" (which is morphology) or iii) Growth of a population (proliferation)
it is important to disambiguate these
I agree I think it would be better to change the namespace to 'single species' phenotypes. @jseager7 is looking into how this would affect PHI-Canto ontology loading and the changes that will need to be made.
It is possible that the chemistry phenotypes will also fall under growth phenotypes.
Regarding the stature phenotype - I'm not sure if this would be plant specific. Perhaps a host mutation in a mouse would alter its height but I'm not sure if this phenotype would be captured more as 'small'? Something to think about when we trial these types of papers for curation.
I also wondered about the 'stature' overlap when describing 'decreased aerial height of filamentous growth'
should this be a child of 'stature' or 'filamentous growth'?
Regarding the 'yeast-like growth form phenotype'. I was curating PMID: 29020037
Zymoseptoria tritici is a dimorphic fungus and is able to growth in both a 'yeast-like' and 'filamentous' form depending on conditions.
The ZtGT2 mutant (grown under various conditions) resulted in normal yeast like growth but abnormal filamentous growth (short filaments, sinusoidal (wavy) filaments) -would this be classed as morphology?
agree I think it would be better to change the namespace to 'single species' phenotypes.
Sorry that I only just spotted that. It's amazing that you can become so accustomed to using a phrase and then suddenly realise that it's wrong !
Yes, the chemistry should probably be under growth phenotypes if they are growth assays.
I'm not sure that I would worry about linking these developmental or anatomical concepts in plants and animals at least at the start. I take your point that stature is not unique to plants and that there might be something in common (based on nutrient avaiability), but in general palnt and animal anatomical phenotypes will be completely disjoint. This is different than, say, different parts of vertebrates and flies which map to each other, or all of the cell level stuff which is often the same.
I envisage that eventually if large numbers of terms are required, you will probably end up using external ontologies to describe gross anatomy phenotypes of hosts. We don't want to develop competing ontologies, but we also need to establish which ontology subsets we might require. This approach seems that it will work for this funding period.
Regarding the 'yeast-like growth form phenotype'. I was curating PMID: 29020037 Zymoseptoria tritici is a dimorphic fungus and is able to growth in both a 'yeast-like' and 'filamentous' form depending on conditions. The ZtGT2 mutant (grown under various conditions) resulted in normal yeast like growth but abnormal filamentous growth (short filaments, sinusoidal (wavy) filaments) -would this be classed as morphology?
Yes this one is morphology. Yikes this is complicated! really it now required a life-cycle stage ontology too- but we will fudge that into PHIPO for now!
we need some way of saying
normal yeast-stage morphology
abnormal morphology during filamentous growth
(here the word 'growth' is being used to describe a 'life cycle phase i.e period of growth' rather than a "size increase'...size does increase- but it isn't directly related to the observation, size and shape are independent of each other- you are lucky we have thought about this a lot, although Midori could likely explain it Much better)
just to note we plan to add chemistry phenotypes #43 and spore germination phenotypes #44 under growth phenotypes
I expect the 'penetration phenotypes' #32 (in both single species and pathogen host interaction namespaces) will also fall under growth phenotypes.
In the pathogen host interaction namespace would they also need to link to virulence and pathogenicity phenotypes or is this orthogonal information? One annotation could curate whether the pathogen penetration of the host was normal/abnormal and then a second annotation could indicate how this affected pathogenicity or virulence.
https://www.ebi.ac.uk/QuickGO/term/GO:0036267 might be useful for the logical def
Val
Trying to figure out which parents 'organism growth phenotypes' would have #42
Also @jseager7 where did we get to with changing the namespace name - is there a ticket for this?
I agree I think it would be better to change the namespace to 'single species' phenotypes. @jseager7 is looking into how this would affect PHI-Canto ontology loading and the changes that will need to be made.
where did we get to with changing the namespace name - is there a ticket for this?
@CuzickA There isn't any ticket open for this, and there hasn't been any progress with it, since we still haven't fixed the bugs that occurred in Canto from the last namespace merge. Given recent discussion about the possibility of moving all the terms into one default namespace (and handling term groupings with subsets instead), we might want to wait until we've reached a decision on that before changing the existing namespace again.
Closing this older ticket
Under the single 'organism phenotype namespace' I have begun to merge pathogen and host only growth phenotypes. This is what we currently have![image](https://user-images.githubusercontent.com/11753634/52555137-c7d41500-2de0-11e9-965f-f1433f22da07.png)
Queries 1) what else fits in here? 2) although aiming for species neutral phenotypes we will also need some species specific phenotypes such as filamentous growth -does this seem ok? 3) should reproductive phenotypes be added here or kept separate?