Collecting examples of phenotypes that explicitly refer to stages

[matentzn]

We need to collect a few current and anticipated needs of representing stages directly as part of our phenotype definitions, for example, phenotypes where prenatal time stages are affected. It would be great if you could all give a few examples of phenotypes that mention stages explicitly as part of their logical definition, and three words why it is necessary for the stage to be mentioned. From that, we will try to create a taxonomy of cases where stages are necessary, and then start developing the respective patterns. @sbello @ybradford @dosumis @nicolevasilevsky @LCCarmody @srobb1 @mah11 @chris-grove @Clare72 @pnrobinson

[mah11]

Fission yeast has very few life cycle stages, but we do have phenotypes specific to each:

• vegetative growth phase
• meiotic cell cycle / meiosis & sporulation
• cell quiescence / G0

Cell morphology, a subcellular structure, or a process could differ from wild type in one stage but not the others. If cell cycle phases count as "stages" for this purpose, that brings in oodles more, with similar (proto-)patterns.

A few examples, with partial logical definitions (because I haven't yet learned how to do all of the differentiae):

• FYPO:0002944 ! decreased RNA level during G0

  'has_part' some ('decreased amount' and ('inheres_in' some 'RNA') and ('has_modifier' some 'abnormal') and [whatever would be correct to represent "during G0", referring to GO:0044838])

• FYPO:0004242 ! decreased protein kinase activity during mitotic S phase

  'has_part' some ('decreased rate' and ('inheres_in' some 'protein kinase activity') and ('happens during' some 'mitotic S phase') and ('has_modifier' some 'abnormal'))

• FYPO:0000062 ! abnormal nuclear morphology during vegetative growth (note: refers to GO:0072690)

  'has_part' some ('morphology' and ('inheres_in' some 'nucleus') and ('during' some 'single-celled organism vegetative growth phase') and ('has_modifier' some 'abnormal'))

Is that the sort of thing you're after?

[srengel]

We have similar for budding yeast in APO, but not sure if that’s what you’re going for here.

[chris-grove]

Here's a list (likely not comprehensive) of 246 phenotype terms from the worm phenotype ontology that refer to life stages:

https://www.dropbox.com/s/fq5zav2si71b3os/Life_stage_phenotypes.txt?dl=0

Some examples:

WBPhenotype:0000035 | larval body morphology variant WBPhenotype:0000055 | early larval arrest WBPhenotype:0000163 | clear larva WBPhenotype:0000753 | L3 larval development variant WBPhenotype:0000759 | spindle defective early emb (during early embryo stage) WBPhenotype:0001020 | embryonic lethal late emb (during late embryo stage) WBPhenotype:0001176 | one cell shape defective early emb (one cell embryo shape defect)

Similarly to what @mah11 wrote, these phenotypes are particular to that life stage as in they occur during a particular life stage but not necessarily during other life stages. These are important distinctions in the C. elegans field. I have not written logical definitions for these terms myself, but the constructs that @cmungall had used when proposing EQs include:

'has part' some (during some 'stage') 'has part' some ('inheres in' some 'stage') 'has part' some ('ends_during_or_before' some 'stage') 'has part' some ('starts_during_or_after' some 'stage')

Some constructs refer to a GO term that embeds the life stage like:

'has part' some ('inheres in' some 'process')

where the 'process' is "larval development" or something similar.

I suppose I would think we would want something like:

'has part' some ('occurs during' some 'stage')

or possibly

'has part' some ('inheres in' some 'stage')

[LCCarmody]

This is a bit hard to answer for HP. I'll look deeper, but there are a lot of 'developmental X' terms that describe phenotypes that occur in early stages, usually childhood or 'fetal X'. Very few actually have logical defs. Ex, HP_0001791 Fetal ascites, log def. 'has part' some ('increased amount' and ('inheres in' some ('interstitial fluid' and ('located in' some ('peritoneal cavity' and ('part of' some 'late embryo'))))) and ('has modifier' some abnormal))

HP_0001789 Hyrdops fetalis 'has part' some (edematous and ('inheres in' some embryo) and ('has modifier' some abnormal))

HP_0001197 Abnormality of prenatal development or birth 'has part' some (quality and ('inheres in' some 'late embryo') and ('has modifier' some abnormal))

Not trying to argue these are the best examples, but there are many, many terms, but these are the very few that have log defs. I definitely think there needs to be a distinction between prenatal/fetal and postnatal. There are many terms using developmental, juvenile, or childhood in labels and text defs, but those would be more difficult to develop log. defs.

Adolescent seems to be covered more or less by using puberty to define it:

HP_0000823 Delayed puberty 'has part' some (delayed and ('inheres in' some 'development of secondary sexual characteristics') and ('has modifier' some abnormal))

HP_0008185 precocious puberty in males 'has part' some (premature and ('inheres in' some 'development of secondary male sexual characteristics') and ('has modifier' some abnormal))

[dosumis]

Use cases:

Some general questions worth asking about this:

• Can your use case be driven via classes in referenced ontologies which are partially defined in terms of stage. e.g. DPO needs larval and adult behavioral phenotypes to be distinct, but this can work by making grouping terms for larval and adult behavior and referencing these in DPO behavioral phenotype classes.
• Do you require some form of temporal reasoning, and if so:
  • does it only involve reasoning about one stage being part of (encompassed) by another, e.g. L3 part of larval stage => L3 larval development variant is_a larval developmental varient
  • or does it involve being able to group all stages before or after some specified stage? This is possible (with some exceptions), but needs carefuls axiomatization.

DPO

DPO stages of lethality terms refer directly to stasges (pattern is quite complex, uses some simple temporal reasoning). We will soon need reference for larval vs adult behavioral phenotypes, which will be very useful to segregate - although we could probably do that on the behavior terms.

[pfey03]

sorry for adding to this late.

developmental stages Dicty. e.g.

• vegetative growth
• streaming
• aggregation
• slug migration
• culmination
• sporulation

[matentzn]

Also when working on this, take a look at @pnrobinson use case here:

https://github.com/obophenotype/human-phenotype-ontology/issues/7079#issuecomment-936549133

[pfey03]

DDPHENO:0000026 | development arrests at mound stage | DDDEV:0000002 | mound stage DDPHENO:0000155 | development arrests at tipped mound stage | DDDEV:0000004 | tipped mound stage DDPHENO:0000159 | development arrests at loose mound stage | DDDEV:0000003 | loose mound stage DDPHENO:0000168 | development arrests at slug stage | DDDEV:0000006 | slug stage DDPHENO:0000685 | development arrests at standing slug stage | DDDEV:0000005 | standing slug stage DDPHENO:0000686 | development arrests at early culminant stage | DDDEV:0000007 | early culminant stage DDPHENO:0000687 | development arrests at mid culminant stage | DDDEV:0000008 | mid culminant stage DDPHENO:0000688 | development arrests at late culminant stage | DDDEV:0000009 | late culminant stage

Parent: DDPHENO:0000154 ! aberrant sorocarp development

[sbello]

About to add a term for developmental cataracts https://github.com/mgijax/mammalian-phenotype-ontology/issues/3499 matching term is present in HPO