Open ncatlett opened 6 years ago
Does it make sense to associate units separately from the values that are being measured? I had written BEL like this before:
SET Citation = {"PubMed", "23484434"}
############
# Figure 4 #
############
# Disambiguation: TAGI is short for "tau aggregation inhibitors"
SET Evidence = "Redox cycling of methylene blue and TAGI potencies. Upper
panel (black): Reduced leuco form 2 and structural overlap with inactive
phenothiazines 3, 4, and 5. Lower panel (blue): Structural overlap between
oxidized form 1 and azure A (6), reduced anthraquinone, 8, and 9, showing
the extended π-conjugated system (ND: not determined)."
SET Page = "D"
SET Figure = "4"
# Disambiguation: CHEBI:22580 label "anthraquinone"
# Disambiguation: HBP00006 label "Tau aggregates"
SET ArticleCompoundNumber = "7"
SET MeasurementType = "IC50"
SET MeasurementUnits = "μM"
SET MeasurementRelation = "~"
SET MeasurementValue = "2"
a(CHEBI:"22580") -| a(HBP:HBP00006)
UNSET {ArticleCompoundNumber, MeasurementType, MeasurementUnits, \
MeasurementRelation, MeasurementValue}
Or what about when there's a numerical range, like in:
# Disambiguation: CHEBI:6872 label "methylene blue"
# Disambiguation: HBP00006 label "Tau aggregates"
SET ArticleCompoundNumber = "1"
SET MeasurementType = "IC50"
SET MeasurementUnits = "μM"
SET MeasurementRelation = "="
SET MeasurementRangeLower = "1.9"
SET MeasurementRangeUpper = "31"
a(CHEBI:"6872") -| a(HBP:HBP00006)
UNSET {ArticleCompoundNumber, MeasurementType, MeasurementUnits, \
MeasurementRelation, MeasurementRangeLower, MeasurementRangeUpper}
How would measurement units be specified where there are multiple values being specified? E.g., there could be numeric annotations associated with the dose of a drug or ligand used in a treatment, as well as the change in an output variable (fold change, amount of cytokine secreted, etc.)
I would not separate the units from a numeric annotation. Ideally, the units would be included as part of the annotation, e.g., "2 hours"
For a given annotation, is it a namespace annotation, or a numeric one? Could we also have more complicated objects with exotic schemata?
Numeric annotations - BEP7
closes #15