Acidosis & Alkalosis logical definition causes lots of bad inference

[dosumis]

HPO term

acidosis: 'has part' some ('increased amount' and ('characteristic of' some ('chemical entity' and ('part of' some 'bodily fluid') and ('has role' some acid))) and ('has modifier' some abnormal))

Leads to classification of increased levels of substances in bodily fluids that recorded as acids by CHEBI e.g.- Hyperglutaminemia is an inferred subclass due to it being glutamine being an amino acids (it's actually also inferred to be Akalosis as well due. 'has part' some ('increased amount' and ('characteristic of' some (glutamine and ('part of' some blood))) and ('has modifier' some abnormal))

Suggested revision of logical definition

The definition should not be based on the nature of chemicals but on abnormal pH of the bodily fluid =>

'has part' some ('increased acidity' and ('characteristic of' some 'body fluid') and (has_modifier some 'abnormal')

ditto Alkalosis.

[dosumis]

(CC @rays22 )

[pnrobinson]

One of the main difficulties is that what we actually observe is acidemia (this is a pH) but what is often recorded in medical literature is an inferrence (acidosis, which is sometimes used interchangably with acidemia). There are then subclasses that reflect the specific process causing acidemia. While this is often pretty accurate, I think that it is confusing to have separate terms in the HPO code acidemia and acidosis and we should consider restructuring.

[LCCarmody]

I would only add the logical defs. I will leave the restructuring up to you.

[LCCarmody]

Following up on this, if acidosis is : 'has part' some ('increased acidity' and ('characteristic of' some 'body fluid') and (has_modifier some 'abnormal')

Then what should 'lactic acidosis' be?

[pnrobinson]

@matentzn I am not sure we should worry that much about these definitions, and maybe we can close this? I do not know what the open action item is?

[matentzn]

The "lazy" way out to satisfy the request is to delete the acidosis and alkalosis logical definitions. The right way would be to ask Leigh to discuss this with @rays22 in the next uPheno call and fix it.

You could also to both:

1. Delete now
2. Close issue
3. Ask Leigh and Ray to add a corrected EQ if they think it is important

[pnrobinson]

Let's leave this open until @LCCarmody and @rays22 can deal with it!