Review: increasedLevel

[nicolevasilevsky]

increasedLevel.yaml

could we rename this to increasedAmount?

[matentzn]

I find this pattern highly unspecific. I would like to know what it pertains to.. High amounts of what?

[nicolevasilevsky]

In HPO, it is used for increased levels of chemicals, cell counts, protein levels, biological processes (which doesn't really seem correct to me), and anatomical entities (which also seems odd)

[matentzn]

Yeah, that seems too wide a scope. Increased level is imho not the same as increased number of.. I think this pattern should apply to chemicals only.. @ybradford @sbello?

[sbello]

This came up in https://github.com/obophenotype/upheno/issues/327 as well. Yes, the desire is to split level of chemical from amount of physical entity (cells, tissues). We can either repurpose the non-specific level patterns or obsolete these and develop new ones.

[ybradford]

@matentzn Level to me implies chemical - or some measurable substance. I would not use it for increased number of cells, increased number of parts (anatomical), - this should use increased number of and for process - this should use increased process quality or extra processual parts

[matentzn]

Okay, lets restrict the "level" patterns to chemicals and develop has increased number of patterns for the anatomical entities. Regarding increased "level" of process equivalents: rather than extra processual parts, which suggest to me that there is something qualitatively different going in the the process (something extra), perhaps increased rate is a better choice? @ybradford @sbello @nicolevasilevsky

[mah11]

Regarding increased "level" of process equivalents: rather than extra processual parts, which suggest to me that there is something qualitatively different going in the the process (something extra), perhaps increased rate is a better choice?

In FYPO, we have found it Very Useful Indeed to distinguish increased/decreased frequency or extent of occurrence of a process from increased/decreased "rate" in the sense of how fast it goes/how long it takes. We've been using the PATO 'occurrence' terms PATO:0002051/PATO:0002052 for the former and 'rate' PATO:0000911/PATO:0000912 for the latter. Happy to convert our design patterns if that's not accurate, but in any case we want to be able to continue to make the distinction between speed and occurrence.

I agree that neither 'increased rate' nor 'increased occurrence' looks like it means the same thing as 'extra processual parts'.

[matentzn]

@sbello @nicolevasilevsky @ybradford What do you think about @mah11 suggestion? should we have increasedRateOfProcess and increasedFrequencyOfProcess with the semantics Midori is suggesting?

[dosumis]

From discussion in meeting:

level vs amount in terrm names (do we even need to standardize?)

• amount is favoured by ZP editors over levels in order to prevent churn, but we need (abnormal/increased/decreased levels as synonym). ZP term names are automated so PATO term choice determines default term names here (unless over-ridden).
• 'levels' is currently used by MP and HP

Increased rate vs increased frequency:

PATO has rate as parent. This seems reaosonable for many use cases, but we should discuss Midori's use cases at the workshop.

[matentzn]

@nicolevasilevsky can you tell me how you distinguish increased level from increased concentration terms in HPO? increasedLevelOfChemicalEntity

[nicolevasilevsky]

I think increasedLevelOfChemicalEntity was an old name, which has now been replaced with increasedConcentrationOfChemicalEntity

[matentzn]

Make sure this is true :) because if I am not mistaken they are both there.

[nicolevasilevsky]

There is abnormallyIncreased and Decreased Level but not abnormal level

[nicolevasilevsky]

So I think for terms like this there is: abnormalConcentrationOfChemicalEntity abnormalConcentrationOfChemicalEntityInLocation

abnormallyDecreasedLevelOfChemicalEntity abnormallyDecreasedLevelOfChemicalEntityInLocation

abnormallyIncreasedLevelOfChemicalEntity abnormallyIncreasedLevelOfChemicalEntityInLocation

[nicolevasilevsky]

HPO currently uses 'amount' instead of 'concentration' for terms like HP_0045056 'Abnormal levels of alpha-fetoprotein'

[pnrobinson]

There are important differences between amount and concentration, and for instance in urine the concentration does not mean much, it is the amount in 24h. (Concentration in urine depends on factors like how much coffee you drink, but the amount in 24h does not). So please keep both. We may need to revise some of the definitions if this is not clear.

[nicolevasilevsky]

Thanks @pnrobinson!

@matentzn it sounds like we need both patterns, yes? I can create the additional patterns we need, please let me know if that sounds good

[matentzn]

@nicolevasilevsky Could you draft two detailed human-readable definitions that clarify exactly what the difference is between the two and update the two patterns accordingly; I dont quite understand it still, but I dont have too :P Maybe just add a comment here saying:

abnormalConcentrationOfChemicalEntity: "The amount of X per Y (or whatever)" abnormalLevelOfChemicalEntity: "..."

We will have the community review this afterwards.

[pfey03]

@nicolevasilevsky @matentzn Coming late into this, in Dicty we have level for chemical entities. For example have this phenotype increased intracellular glucose level Def.: Increase in amount or accumulation of cytosolic glucose. Despite the word 'amount' in the def, I would not use it here. abnormalLevelOfChemicalEntity "the measured accumulation of a chemical" or such

[matentzn]

@nicolevasilevsky :P Sorry to be annoying; Just want to close this issue down soon; Thanks!

[nicolevasilevsky]

Not sure if I can properly define the difference between concentration and level, but I will try.

According to wikipedia, concentration is the abundance of a constituent divided by the total volume of a mixture.

HPO only uses 'increased concentration' in one term: HP_0002910 'Elevated hepatic transaminase' 'has part' some ('increased concentration' and ('inheres in' some (protein and ('capable of' some 'transaminase activity') and ('produced by' some liver))) and ('has modifier' some abnormal)) I don't think we have a pattern for this

For abnormal/increased/decreased chemicals, the HPO is using the pattern abnormallyDecreasedLevelOfChemicalEntity (or increased)

For the purposes of HPO, I think we need these patterns:

abnormalLevelofChemicalEntity abnormalLevelofChemicalEntityInLocation (which would use the quality 'amount') we would need to add these

abnormallyDecreasedLevelOfChemicalEntity abnormallyDecreasedLevelOfChemicalEntityInLocation

abnormallyIncreasedLevelOfChemicalEntity abnormallyIncreasedLevelOfChemicalEntityInLocation these already exist

[nicolevasilevsky]

Potential human readable defs are below:

abnormalLevelOfChemicalEntity Deviation from normal level of a chemical entity or protein.

abnormallyDecreasedLevelOfChemicalEntity Reduction in the level of a chemical entity or protein.

abnormallyIncreasedLevelOfChemicalEntity Elevation in the level of a chemical entity or protein.

[matentzn]

abnormalLevelofChemicalEntity, abnormalLevelofChemicalEntityInLocation (which would use the quality 'amount') we would need to add these

(these are in the dosdp-patterns folder)

[nicolevasilevsky]

Ah, thanks @matentzn!

[LCCarmody]

So...looks like this ticket was never quite resolved? We have the same issue here. Basically, we need an abnormalLevelofChemicalEntity and/or abnormallyIncreasedConcentrationofChemicalEntity/abnormallyDecreasedConcentrationofChemicalEntity.

I proposed in the HP ticket that we use abnormalLevelofChemicalEntity for Abnormal circulating cholesterol concentration to fit with the children of the term and most of HP terms. Should I start to make this pattern?