Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Pattern build issue with FoodON/CHEBI imports #230

Closed
LCCarmody opened this issue May 28, 2021 · 10 comments
Closed

Pattern build issue with FoodON/CHEBI imports #230

LCCarmody opened this issue May 28, 2021 · 10 comments

Comments

@LCCarmody
Copy link
Member

This has only been seen on my new branch (that I am aware of) #229 so you would have to look on this branch to see what I'm talking about.
Basically, I had to update a few terms in the nutrition_supplementation_foodon pattern.

First, it appeared that FoodOn had obsoleted their Vitamin C and Vitamin D supplement terms, so I needed to update those. FoodOn is now importing Vitamin C and Vitamin D from CHEBI, so I updated those terms in my patterns.

However, now, I am having issues with OTHER terms imported from Foodon. Specifically, MAXO_0000761 which should be vitamin B supplementation, but is now 'B vitaminvitamin B supplementation'.
MAXO_0000108 iron supplementation, but now synonyms are 'iron atomiron supplemental intake'
MAXO_0010100 dietary iron intake avoidance now has synonyms that are 'avoid iron atomiron intake'
MAXO_0010096 dietary sodium intake avoidance now has synonyms that are 'sodium atomsodium consumption avoidance'
MAXO_0010115 dietary calcium avoidance now has synonyms that are 'avoid calcium atomcalcium intake'

I believe the source of the issue is that there are more than one rdfs:label for the terms I am pulling from Foodon (that were imported from CHEBI) and they are not identical. Somehow, when robot pulls these terms from more than one rdfs:label and mashes them together. Unclear why this is happening.

I can write patterns to work around this, but I thought I would point it out. Not sure if this is an issue with FOODON importing the CHEBI term and then altering the label or if it is an issue with Robot. Or both.
@matentzn @ddooley

Anyways, this is not critical to be solved this minute--as this is a side branch I've been working on. And, as I said, I can come up with a workaround pretty easily, but I wanted to point it out. Thanks!

@ddooley
Copy link

ddooley commented May 29, 2021

So I am sure this is related to FoodOn adopting CDNO for its nutritional component hierarchy, and using CDNO's organization of CHEBI terms (and deprecating in favour of them) where possible. At moment CDNO is requesting some changes in CHEBI re. vitamin role / material entity, and that is being sorted out. However I'll have a look to see if there are duplicate labels somehow directly in FoodOn, upstream of your MAXO terms. FoodOn would never alter rdfs:labels intentionally! Pretty sure CDNO wouldn't either but we'll see...

d.

@ddooley
Copy link

ddooley commented May 29, 2021

So I'm gathering that this is about DOSP patterns that need to be adjusted to include 'term replaced by' ids? In FoodOn, the switch is to CHEBI terms directly. However, for Vitamins in particular, there is some confusion at moment about CHEBI Vitamins, which should be resolved fairly soon, see thread: CompositionalDietaryNutritionOntology/cdno#57

@kaiiam
Copy link

kaiiam commented Jun 1, 2021

@ddooley is this due to those CHEBI vitamin terms being roles rather than material entities? Aka were the original (now deprecated) foodon vitamin terms material entities?

@LCCarmody The long and short of the CDNO issue above is that CHEBI vitamins are a mix of roles and chemical entities so @LilyAndres is working with the CHEBI curation team to sort that out. Hopefully that will help resolve the issue at hand here.

@LCCarmody
Copy link
Member Author

Thanks for getting back to me. It seems somewhere along the way there re 2 different 'rdfs:labels' instead of using a synonym. For some reason, when I do a build, it combines the 2 different labels in an odd way. I can override this, but wherever there is a change in label name, that needs to be changed. Adding a screenshot of how CHEBI appears in my import...3 rdfs:labels are 'calcium atom', but the first one is 'calcium'. I'm not sure why ROBOT combines the 2, but it does.
Screen Shot 2021-06-01 at 9 14 29 AM

Thank you for looking into this.

@ddooley
Copy link

ddooley commented Jun 1, 2021

I see now this discrepancy is coming in via CDNO. I will raise an issue there. This is a screenshot of CDNO calcium, which comes in via CHEBI.

image

Reported in CompositionalDietaryNutritionOntology/cdno#64

@LCCarmody
Copy link
Member Author

Thanks @ddooley

@kaiiam
Copy link

kaiiam commented Jun 2, 2021

Thanks all will respond in CompositionalDietaryNutritionOntology/cdno#64

@LilyAndres
Copy link

Thanks for this, please see my reply in Southern-Cross-Plant-Science/cdno#64

@ddooley
Copy link

ddooley commented Jun 16, 2021

The update is that CHEBI will be implementing both a vitamin role and vitamin molecule (vitamer) branch in the coming weeks. CDNO and thus FoodOn will then be able to make appropriate logically consistent references.

@LCCarmody
Copy link
Member Author

Not sure when this was fixed, but all seems to be good. Thanks so much @ddooley and @LilyAndres

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants