New Term Request: Adding isAnatomicalEntityFor and more.

[BideZ]

Hello Everyone! I am here submitting several relations to the RO including:

isAnatomicalEntityFor

• def:A relation describing an anatomical feature of the referenced object with a material entity.

• use case: 'anterior frontopolar cortex' isAnatomicalEntityFor some epilepsy

isBrainConnectivityFor

• def: The subject describes the techniques used for estimating relations between brain regions in a given condition (object).

• use case: Electrocorticography isBrainConnectivityFor some epilepsy

isCellularProcessFor

• def: The subject describes any process that is carried out at the cellular level, but not necessarily restricted to a single cell for the object."

• use case: 'long term potentiation' isCellularProcessFor some epilepsy

isClassificationFor

• def: The subject is a subcategory of the object with certain criteria.

• use case: 'focal epilepsy' isClassificationFor some epilepsy

isDiagnosisFor

• def: The subject is a technique carried out in a clinical assessment for evaluation of a condition (object).

• use case:electroencephalography isDiagnosisFor some epilepsy

isEtiologyFor

• def: The subject is a cause or result in the pathogenesis of a condition (object).
• use case: 'Gene Abnormality' isEtiologyFor some epilepsy

isImitatorFor

• def: The subject shares some similarities with a certain condition (object) and can be misdiagnosed as this condition.

• use case: 'Eidetic Imagery' isImitatorFor some epilepsy

isRiskFactorFor

• def: The subject possesses characteristics that lead to an increased risk in resulting in a certain phenotype or condition (object).

• use case: 'injury' isRiskFactorFor some epilepsy

isSeizureClassificationFor

• def: The subject assignes the object a seizure class.

• use case: 'negative motor seizure' isSeizureClassificationFor some epilepsy

isSignAndSymptomFor

• def: The subject describes a sign/symptom for the object.

• use case: 'aphasia' isSignAndSymptomFor some epilepsy

isSyndromeFor

• def: The subject describes a syndrome correlated to a condition (object).

• use case: 'epilepsy with generalized tonic-clonic seizures' isSyndromeFor some epilepsy

isTreatmentFor

• def: The subject describes a clinically approved drug or therapy for a condition (object).

• use case: 'Speech Therapy' isTreatmentFor some epilepsy

isEndpointFor

• def: The subject whose level or measurement confirms an onset or an existance of a condition (object).

• use case: 'Mini-Mental State Exam score' isEndpointFor some 'Alzheimer's disease'

isModelFor

• def The subject serves as an in vitro/in vivo model for the study of a certain phenotype (object) during pre-clinical phase.

• use case: 'Neuro-2a cell' isModelFor some 'Alzheimer's disease'

We are currently using these terms for text mining bins and they are applied in many ontologies constructed by our institute. As suggested by @matentzn, we thought we could contribute to the community by adding these terms into the RO, so the other ontologies we construct in the future can be aligned. Please tell us how you think about these ones!

We are looking forward to your feedbacks!

[cmungall]

Hi @BideZ, my first recommendation would be to look at what exists in RO already and see what fits your needs. For example, there should already be relations for model-of and treats. Did you come across these and find them unsatisfactory?

I recommend you also read our guide https://oborel.github.io/obo-relations/ (it is imperfect and a work in progress, suggestions for improvements always welcome).

It looks like you are looking for relations for use with concepts that are present in existing OBO ontologies such as phenotype ontologies or disease ontologies. Are these relations to go into these ontologies (HP, MP, Mondo, DOID, etc)? You might want to look at these ontologies and see how they are modeling some of the same kinds of statements. For example, these ontologies already have linkages to GO biological processes for concepts like epilepsy. I would strongly advise working with these groups rather than coming up with alternative ways to say what is already being said.

isClassificationFor: can you explain why you can't just use subClassOf?

[BideZ]

@cmungall Thank you very much for your feedback! We are aiming at using these object properties for generation of bins used in text mining (as shown in the picture).

We have been seeking relations that suit the most to our applications in many ontologies such as RO, HP, SYMP and etc. Most of the terms found in these ontologies might be applicable, yet they are either too broad in definitions or restrained to a specific type of classes. It might create ambiguity when applying these terms in our scope of study.

Taking 'is classification for' as an example: the most equivalent relation that we found there is 'disease has feature' from RO. It is defined as 'A relationship between a disease and some feature of that disease, where the feature is either a phenotype or an isolated disease.' The relations are comparable to some extent. However, 'is classification for' stands for a particular categorization stricted defined through clinical markers of a condition (in Alzheimer's disease / Epilepsy).

Since we are using these object properties with comparably more detailed definition, we could hardly find relations from the underlying ontologies that tell the same story.

We are looking forward to your feedbacks again!

[BideZ]

Hi Everyone. After several rounds of discussion with one of the reviewers for ADO ontology, we have changed the name and definition of several relations as described below:

• [amendment] is risk factor For: The subject possesses characteristics leading to an increased risk in developing a certain phenotype or condition (object). The example of risk factors could be age, family history, consumption of certain products, being exposed to chemicals or radiations, having certain underlying conditions or being infected by an microorganism

• [addition] Variant is Genetic Risk Factor For: The subject gene has variation that can lead to an increased disease risk, but does not directly cause a disease.

Note: the original relation 'is risk factor for' is now separated into 'is risk factor for'(non-genetic factors) and 'variant is genetic risk factor for' (genetic factors).

• [amendment] is Clinical Marker For(originally 'is Endpoint For'): A subject that is a symptom, a sign, an underlying disease, a molecular abnormality, or a clinical result/outcome characterised as the key marker for the diagnosis/onset/classification of a certain condition (object).

[nlharris]

So is this still a new term request?

[nlharris]

Can this be closed, or is there still work to be done here?

[wdduncan]

@BideZ
Thank you again for your submission. After reviewing and discussing this with the larger RO group, we do not think that these relationships can be admitted to RO. RO is supposed to provide a limited list of relationships that can be re-used across OBO ontologies. While the presence of application ontologies in OBO that have a restricted scope and idiosyncratic classes can be fine, 'application relationships' are not. Other users of RO will come across your relationship labels, and will not understand how they relate to existing RO relationships, thus causing confusion. For your planned application of using these for text mining, you are of course free to use these, but we cannot include them in RO which has a more restricted definition of what relationships mean.