Increased/decreased chemical concentration in location

[seger]

New patterns requested by Xenbase curators.

[matentzn]

This pattern has been discussed a few times, since http://purl.obolibrary.org/obo/upheno/patterns-dev/abnormalConcentrationOfChemicalEntity.yaml exists, we will admit it, but I think if we really want take the opportunity here to very clearly define under which exact circumstances we should use "abnormal level of" vs "abnormal concentration of".

[seger]

@matentzn The Xenbase curators say that "abnormal level of" suffices instead of "abnormal concentration of", so we can withdraw this request.

[rays22]

@matentzn The Xenbase curators say that "abnormal level of" suffices instead of "abnormal concentration of", so we can withdraw this request.

@seger , @matentzn
"Level" is very ambiguously when people describe phenotypes. It may mean "amount", "total amount", or "concentration" depending on the context.

PATO:0001163 decreased concentration is a subclass of PATO:0001997 decreased amount so we can stop here.
However, there are valid cases when "decreased concentration" would be more appropriate to describe a phenotype than "decreased amount". For example, when "decreased amount" is intended to mean "decreased total amount". In those case "decreased concentration" does not necessarily lead to "decreased total amount". OK, we do not have "total amount" (as a subclass of "amount") or a "decreased total amount" (as a subclass of "decreased amount") in PATO yet, but I think we should to distinguish "decreased total amount" and "decreased concentration" phenotypes.

Similar reasoning should apply to "increased amount" and "increased concentration".

I have no objections adding these templates if they are needed for data annotation.

Let me know if you want me to proceed with the review.

[seger]

@rays22 @matentzn I'm in favor of adding the templates. That way, we can capture what the curators originally intended with those annotations. Thank you.

[sbello]

The reason we had backed off of creating concentration and level patterns was that when curators went back and reviewed the annotations we found authors often used these terms interchangeably or at the very least inconsistently. We decided to use level as it was the broader term. Having both patterns could be confusing and lead to the same phenotype being curated differently by different curators. Sue

…

On Wed, Feb 5, 2025 at 7:49 PM Erik Segerdell ***@***.***> wrote: @rays22 <https://github.com/rays22> @matentzn <https://github.com/matentzn> I'm in favor of adding the templates. That way, we can capture what the curators originally intended with those annotations. Thank you. — Reply to this email directly, view it on GitHub <#973 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACTYTULUFR77BHUUT6ZHFLT2OKWQNAVCNFSM6AAAAABVM3HC26VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMZYGM2TQMJXGA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[rays22]

Could you add two or three phenotype examples in the description section of each template?

[rays22]

@seger
Could you please ask the curators for some phenotype examples when using the more general "level"-pattern (as in abnormallyDecreasedLevelOfChemicalEntityInLocation) would miss distinguishing some phenotypes that are based on concentration measurements?

We need to provide examples of justified cases for this template.
It may be confusing that some phenotype categories that use the
abnormallyDecreasedLevelOfChemicalEntityInLocation or the
abnormallyIncreasedLevelOfChemicalEntityInLocation
patterns are based on concentration measurements.

However, as @sbello noted, the term "level" (used as a synonym of PATO "amount") has been intended to include both "concentration" and non-concentration-type amount qualities (like "total amount"). The meaning of "level" in uPheno is removed from the strict primary measurement type (concentration- or other measurement). When the quantitative observation/measurement is interpreted, the original dimensions are lost and, the phenotype is interpreted on a relative scale that includes abnormally decreased or abnormally increased (and compared to a reference).
See for example the EQs for these phenotype terms:

• HP:0002902 Hyponatremia
• MP:0005634 decreased circulating sodium level
• HP:0003228 Hypernatremia
• MP:0005633 increased circulating sodium level

Technically all of these phenotypes are based on concentration measurements. It is not typically feasible to measure (quickly and in a non-invasive way) the total amount of sodium in the blood of mouse and human, so we do not loose any granularity by using the more general "level" pattern (as far as I know).

I think there could be justified cases for the more specific "concentration" patterns, but we need two or three examples in the description of this pattern so that we can help curators decide how and when to use it.

[rays22]

Could you please ask the curators for two or three phenotype examples when using the more general "level"-pattern (as in abnormallyIncreasedLevelOfChemicalEntityInLocation) would miss distinguishing some phenotypes that are based on concentration measurements from other types (e.g. total amount) of measurements?