Role of Propositions
Propositions represent possible facts about the domain of discourse that may be true or false. They are abstract entities that capture a ‘sharable’ piece of meaning whose identity is independent of space and time, and which make no claim as to whether the sentiment it expresses is true or not.
As illustrated in the Data Structures section, the job of a Proposition object is to encapsulate a structured representation of a possible fact so it can be referenced and reused across Statements and Evidence Lines.
Propositions in Statements
In a Statement, a given Proposition may be asserted to be true or false, or assessed to report the strength of existing evidence for or against it (e.g. “there is presently likely evidence disputing the proposition that ‘SOS1:c.3322G>A is causal for RASopathy’”).
An example of a Proposition as an assessed possible fact is illustrated below.
Legend: Left panel (Model). Abridged version of the Variant Pathogenicity Statement model. Center Panel (Data Example). An example of a Variant Pathogenicity Statement object (note use of shorthand syntax to capture values that should be wrapped in MappableConcepts). Right Panel (Meaning). Plain language meaning of what structured data in the example reports to be true. Proposition-related content in each panel is highlighted.
Propositions in Evidence Lines
In an Evidence Line, the targetProposition attribute captures the possible fact toward which Evidence Items are assessed and scored (e.g. that a gnomAD population frequency evidence item is evaluated toward the proposition that “SOS1:c.3322G>A is causal for RASopathy” when assessing the evidence as providing strong support against it). An example of a Proposition* in this role is illustrated below.
Legend: Left panel (Model). Abridged version of a Pathogenicity Evidence Line model. Center Panel (Data Example). An example of a Pathogenicity Evidence Line object (note use of shorthand syntax to capture values that should be wrapped in MappableConcepts). Right Panel (Meaning). Plain language meaning of what structured data in the example reports to be true. Proposition-related content in each panel is highlighted.
In VA-Spec data, Propositions are used only in the context of a Statement or Evidence Line, as they convey no knowledge in the absence of the assessments these objects provide.
A Statement references its Proposition via its proposition attribute, while an Evidence Line references the possible fact its evidence is assessed against via its targetProposition attribute. In practice a Proposition should always be given for a Statement that stands on its own, but targetProposition may be omitted from an Evidence Line when it would simply repeat the Proposition of the Statement it supports. For example, in the data example here the root Statement asserts the same Proposition (Proposition001) toward which its two Evidence Lines evaluate the support provided by population frequency and functional impact data. This Proposition object is explicitly referenced in the Evidence Lines in the example, but omission of this reference is permissible, and would imply that the target Proposition here is the same as that in the root Statement.
For more information, see the Proposition page, and related Design Decision.