fixes to development graph

[gocentral]

In order to proceed with the cross-product work, we need as many of the core primitive terms in GO to be disjoint. Doing this now would result in about 200 violations in the development graph. Many of these are actually mistakes in the ontology, although some are debatable.

In the report below/attached, the first two columns are proposed disjoint pairs, and the 3rd colum is the violating term.

GO:0009653-anatomical structure morphogenesis GO:0008283-cell proliferation GO:0060664-epithelial cell proliferation involved in salivary gland morphogenesis

OBVIOUS: change to part_of SGM

GO:0009653-anatomical structure morphogenesis GO:0016477-cell migration GO:0022007-convergent extension involved in neural plate elongation GO:0009653-anatomical structure morphogenesis GO:0016477-cell migration GO:0060975-cardioblast migration to the midline involved in heart field formation

OBVIOUS: change to part_of

GO:0009653-anatomical structure morphogenesis GO:0030154-cell differentiation GO:0060672-epithelial cell morphogenesis involved in placental branching

OBVIOUS: change to part_of

GO:0048856-anatomical structure development GO:0048646-anatomical structure formation involved in morphogenesis GO:0001410-chlamydospore formation GO:0048856-anatomical structure development GO:0048646-anatomical structure formation involved in morphogenesis GO:0030437-ascospore formation

remove isa paths, these are plain formation terms

GO:0048646-anatomical structure formation involved in morphogenesis GO:0071695-anatomical structure maturation GO:0021691-cerebellar Purkinje cell layer maturation

I think the 2nd isa should be removed

GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0003202-endocardial cushion to mesenchymal transition involved in cardiac skeleton development GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0007470-prothoracic disc morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0035114-imaginal disc-derived appendage morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0042246-tissue regeneration GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0048611-embryonic ectodermal digestive tract development GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0060977-coronary vasculature morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0061113-pancreas morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0061197-fungiform papilla morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0061273-mesonephric distal tubule morphogenesis GO:0048856-anatomical structure development GO:0009653-anatomical structure morphogenesis GO:0061301-cerebellum vasculature morphogenesis

all "morphogenesis" terms should isa morphogenesis only; all development terms should be usa development

GO:0048856-anatomical structure development GO:0030154-cell differentiation GO:0001410-chlamydospore formation GO:0048856-anatomical structure development GO:0030154-cell differentiation GO:0003334-keratinocyte development GO:0048856-anatomical structure development GO:0030154-cell differentiation GO:0003398-glial cell differentiation involved in amphid sensory organ development GO:0048856-anatomical structure development GO:0030154-cell differentiation GO:0030437-ascospore formation GO:0048856-anatomical structure development GO:0030154-cell differentiation GO:0042692-muscle cell differentiation

fixes for all these are obvious I think

GO:0009653-anatomical structure morphogenesis GO:0048589-developmental growth GO:0009826-unidimensional cell growth GO:0009653-anatomical structure morphogenesis GO:0048589-developmental growth GO:0042246-tissue regeneration GO:0009653-anatomical structure morphogenesis GO:0048589-developmental growth GO:0048673-collateral sprouting of intact axon in response to injury GO:0009653-anatomical structure morphogenesis GO:0048589-developmental growth GO:0060516-primary prostatic bud elongation GO:0009653-anatomical structure morphogenesis GO:0048589-developmental growth GO:0060523-prostate epithelial cord elongation

TRICKY

GO:0048646-anatomical structure formation involved in morphogenesis GO:0008283-cell proliferation GO:0003263-cardioblast proliferation

I think this should be part_of heart formation

GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0001838-embryonic epithelial tube formation GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0001843-neural tube closure GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0060431-primary lung bud formation GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0060513-prostatic bud formation GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0060615-mammary gland bud formation GO:0048646-anatomical structure formation involved in morphogenesis GO:0009653-anatomical structure morphogenesis GO:0061130-pancreatic bud formation

I think all gross anatomy "formation" terms should be isa GO:0048646 and part_of ASM

GO:0048646-anatomical structure formation involved in morphogenesis GO:0030154-cell differentiation GO:0030220-platelet formation GO:0048646-anatomical structure formation involved in morphogenesis GO:0030154-cell differentiation GO:0030435-sporulation resulting in formation of a cellular spore

not sure about these

GO:0048856-anatomical structure development GO:0048589-developmental growth GO:0042246-tissue regeneration

this one is also under morphogenesis. the text def says "re-growth". but is the text-def correct, the sense GO uses growth?

GO:0001775-cell activation GO:0008283-cell proliferation GO:0046651-lymphocyte proliferation

tricky; proliferation follows activation

GO:0001775-cell activation GO:0030154-cell differentiation GO:0002268-follicular dendritic cell differentiation GO:0001775-cell activation GO:0030154-cell differentiation GO:0002273-plasmacytoid dendritic cell differentiation GO:0001775-cell activation GO:0030154-cell differentiation GO:0030098-lymphocyte differentiation GO:0001775-cell activation GO:0030154-cell differentiation GO:0043011-myeloid dendritic cell differentiation

these are probably fine

Reported by: cmungall

Original Ticket: geneontology/ontology-requests/8114

[gocentral]

Original comment by: cmungall

[gocentral]

Chris,

I think I fixed most of these. Could you run it again?

D

Original comment by: ukemi

[gocentral]

Ping Chris

Original comment by: jl242

[gocentral]

so you don't need me for the most part. I'd recommend putting the disjointFrom statements in the ontology, checking there are no violations in OE, then committing.

More disjoints in GO are great for error checking, and in this section of the ontology are v useful for structuring the logical definitions.

(I've also reduced the priority since I know you're having a sf telethon right now)

Original comment by: cmungall

[gocentral]

• priority: 5 --> 4

Original comment by: cmungall

[gocentral]

This is the tracker item regarding the upper level disjoints for genus terms:

http://sourceforge.net/tracker/index.php?func=detail&aid=3165107&group\_id=36855&atid=440764

here's why this is important for the logical definitions.

In the current xp files we have to use awkward relationships to make sure that the conditions are both necessary and sufficient.

We can let the genus of the definition do more of the work, but we have to ensure there is no "cross-over"

E.g.

BEFORE:

X differentiation = results_in_differentiation_of X X morphogenesis = results_in_morphogenesis_of X

PROPOSED:

X differentiation = cell differentiation and has_target X X morphogenesis = morphogenesis and has_target X

(the actual relation label is not so important for now, the main point is we can use a more generic relation and let the genus do the work)

The problem with implementing this before the disjoints are in place is that deeper down in the ontology graph we had things like GO:0060672-epithelial cell morphogenesis involved in placental branching which had isa paths to both genus terms, so this would have resulted in the reasoner placing this as an isa child of both "epithelial cell differention" and "epithelial cell morphogenesis". there could be even odder results....

so having the disjoints in place guarantees we can safely let the genus do the work in the logical definitions and not get odd inferences

Original comment by: cmungall

[gocentral]

• priority: 4 --> 7

Original comment by: cmungall

[gocentral]

unidimensional cell growth still inherits from both growth and morphogenesis. I think this is fine though, something can be both growing and changing its form.

Although not in the list above, I feel patterning and morphogenesis should be disjoint. This would render the following class unsatisfiable:

[Term] id: GO:0001569 name: patterning of blood vessels namespace: biological_process def: "The process that regulates the coordinated growth and sprouting of blood vessels giving rise to the organized vascular system." [GOC:dph] synonym: "branching involved in blood vessel morphogenesis" EXACT [GOC:dph] is_a: GO:0007389 ! pattern specification process is_a: GO:0048754 ! branching morphogenesis of a tube is_a: GO:0061138 ! morphogenesis of a branching epithelium relationship: part_of GO:0001525 ! angiogenesis

The def says "regulating" but it's not a regulation process.

The is_a relationship to the isa parent "Any developmental process that results in the creation of defined areas or spaces within an organism to which cells respond and eventually are instructed to differentiate." is not immediately clear

Original comment by: cmungall

[ukemi]

@cmungall

I think that we should swap the 'patterning of blood vessels' name with the 'branching involved in blood vessel morphogenesis' synonym and remove it from pattern specification. From the def it looks like this is described as patterning only because it results in the 'pattern' of branches formed. We should keep the patterning term as a synonym because I think it is used to describe the pattern of branches. Is there anything else on this ticket or have the x-products corrected most of this?

-D

[RLovering]

I agree, often the pattern reflects need, although the 'pattern' of the main arteries is probably genetically determined.

Ruth

[ukemi]

OK. I cleaned up the blood vessel terms and their regulation by making the terms an is_a branching morphogenesis and making them part of the morphogenesis of the whole structure. I also did the same for the lymph vessel terms.