BRIDG Mappings for Specimen elements about the anatomical source

[mbrush]

The definition of the ADM.Sample.general_sample_pathology element suggests that the subject of the observation here is the anatomic source of the specimen, not the specimen itself (what type of tissue the specimen was derived/taken from):

"An indicator as to whether a sample is derived from normal or tumor-bearing tissue."

Is there any way in BRIDG to represent the anatomical source/tissue as the subject of the observation here? i.e. can we represent the source tissue in the organism as a specimen, that is separate from the extracted specimen - an observed specimen in situ vs an extracted material specimen in vitro)

And more generally, for all elements classified as being about anatomic source/biological source . . . should the mapping path for these make the subject of observations (and perhaps the root of the Mapping path) be the source material, not the specimen (if we consider the specimen to be the material once taken out of the source)

Here are a list of elements in Sample that I would say are about this anatomical source, rather than the specimen itself:

• biospecimen anatomic site
• biospecimen laterality
• distance to normal tumor
• general sample pathology
• tissue type
• sample type
• sample type id

---

Another general issue this may raise is the distinction between observations about a samples made during collection (while it is still in vivo), as being distinct from observations made after extraction, typically by a pathologist, that may confirm (or refute or extend or qualify) the observation that was made earlier during extraction. This is a common phenomenon, and we need a clear and consistent way to represent it. Being able to represent the anatomic source of a specimen separately from the extracted specimen would be one way to support this (assuming there is a way to capture the derivation relationship between them)

[mbrush]

One possibility for BRIDG supporting this interpretation is to use the BRIDG "BiologicEntityPart" class:

Class: BiologicEntityPart
DEFINITION: A limb, organ, or other portion of a biologic entity.  EXAMPLE(S): the left kidney of a person, a dog's right front paw, a patch of skin on a person's left forearm  OTHER NAME(S):  NOTE(S):
	
Associations:
FROM: BiologicEntityPart 0..* > is part of > 1 BiologicEntity -- DESCRIPTION: Each BiologicEntityPart always is part of one BiologicEntity. Each BiologicEntity might have one or more BiologicEntityPart.  DEFINITION:  EXAMPLE(S):  OTHER NAME(S):  NOTE(S): 

TO: BiologicEntityPart 0..1 < be a function performed by < 0..* ExperimentalUnit -- DESCRIPTION: Each ExperimentalUnit might be a function performed by one BiologicEntityPart.  Each BiologicEntityPart might function as one or more ExperimentalUnit.   DEFINITION:  EXAMPLE(S):  OTHER NAME(S):  NOTE(S): 
	
Attributes:
anatomicSiteCode -- CD -- DEFINITION: A coded value specifying the anatomic location or system of the biologic entity part.  EXAMPLE(S): eye, skin, kidney  OTHER NAME(S):  
	
anatomicSiteLateralityCode -- CD -- DEFINITION: A coded value specifying the side of the body (or a paired organ) where the anatomic site is in a biologic entity part.  EXAMPLE(S): left, right, both  OTHER NAME(S):  
	
quantityRange -- URG<INT.POS> -- DEFINITION: An integer falling within minimum and maximum bounds that specifies how many parts that need to be present.  EXAMPLE(S): 2 Skin patches  OTHER NAME(S):  

Using BiolicEntityPart, the mapping path for all* of these ADM elements that are ostensibly about the anatomical source would look something like this:

• BiologicSpecimen <--beAResultOf-- PerformedSpecimenCollection --beParticipatedInBy--> SpecimenCollectonProtocolSubject --beAFunctionPerformedBy--> BiologicEntityPart <--beAFunctionPerformedBy**-- Subject <--beParticipatedInBy-- PerformedObservation <--isAResultOf-- PerformedClinicalResult.value

  or

• BiologicSpecimen <--beAResultOf-- PerformedSpecimenCollection --beParticipatedInBy--> SpecimenCollectonProtocolSubject --beAFunctionPerformedBy--> Person <--isPartOf-- BiologicEntityPart <--beAFunctionPerformedBy**-- Subject <--beParticipatedInBy-- PerformedObservation <--isAResultOf-- PerformedClinicalResult.value

and

• WHERE PerformedObservation --instantiates--> DefinedObservation.nameCode = "tissue type | sample type | sample type id | general sample pathology | distance to normal tumor "

---

* Exceptions include the biospecimen anatomic site and biospecimen laterality elements, as the Biologic Entity Part class has dedicated/explicit attributes to capture info related to location

** At present BRIDG does not allow BiologiEntityParts to be Subjects of PerformedActivities (only Biologic Entities, and Biologic Entity Groups, and Products, and Specimens) - but I think this should be changed.

[mbrush]

Looking ahead to our harmonized CRDC-H model, we might consider a class representing the anatomical source concepts. Representing the source anatomical entity/tissue as a separate/independent object from the extracted specimen would allow us to talk more clearly about characteristics of observations about the source tissue in the context of the source organism

• e.g. its location in the body - absolute and relative (e.g. peritumoral - relative to a tumor)
• e.g. its pathology as observed in vivo (normal, tumor) . . . distinct from the assessed pathology of the sample (as samples taken from presumed normal tissue may end up being assessed as tumor tissue by a pathologist) . . . capture inconsistencies between where we thought we were collecting and what we actually collected

. . . .and any other elements annotated/categorized as being about the Anatomical source