Reflexive is-a links for TTO terms of metatype 'I'

[Cartik]

Hi Chris,

I found out that reflexive "is a" links are not added for taxa which
are associated with phenotypes. For example, one of the NeXML files
asserts

TTO:1005577 exhibits PATO:
0000052^inheres_in(TAO:0000514) ----[1]
Chrisichthys auratus Shape of
frontal bone

Now, the metatype of this node is changed from 'C' to 'I' in the NODE
table. And the reflexive 'isa*' rule in the OBD reasoner does not
allow reflexive links to be asserted for nodes of the 'I' metatype
viz. nodes like TTO:1005577. Is there a reason why this is so?

I changed this around a little bit to allow reflexive links for nodes
of the 'I' metatype as long as they are associated with a non null
source id.

This is because if I search for phenotypes with Siluriformes (which
subsumes Chrisichthys auratus above) as the search parameter, I can
find [1]. But if I search for phenotypes with Chrisichthys auratus as
the search parameter, I have to use a different query to find [1]
because Chrisichthys auratus is not reflexively mapped to itself
through the is a relation.

Regards,

Cartik

[Jim Balhoff]

Why is the metatype changed from C to I?

[Chris Mungall]

On Jun 23, 2009, at 4:15 PM, Cartik wrote:

>
> Hi Chris,
>
> I found out that reflexive "is a" links are not added for taxa which
> are associated with phenotypes. For example, one of the NeXML files
> asserts
>
> TTO:1005577 exhibits PATO:
> 0000052^inheres_in(TAO:0000514) ----[1]
> Chrisichthys auratus Shape of
> frontal bone
>
> Now, the metatype of this node is changed from 'C' to 'I' in the NODE
> table.

Which node are you talking about?

a) TTO:1005577
or
b) PATO:0000052^inheres_in(TAO:0000514)

The metatype of b should definitely be 'C'

And you're representing taxons as classes, so the metatype of a should
also be 'C'

Looks like a bug somewhere in loading. Do you know when the switch
happened?