The purpose of this BEP is to add the relationship noCorrelation to the BEL Language. This relationship would be used to record cases where a correlative relationship between two entities is investigated, but no significant relationship is observed. The BEL causesNoChange relationship is similar in concept, but requires the assessment of a causal relationship.
Type: Standards Track
BEL includes a relationship causesNoChange for cases where the object term is measured after perturbation of a subject term, but no significant change is observed. BEL does not currently include a parallel relationship for non-causal cases, i.e., where correlation between entities is investigated, but no significant correlation is observed.
The noCorrelation relationship will be used to record cases where a correlative relationship between two entities is investigated, but no significant relationship is observed. This will be particularly useful to capture cases where several analytes are assessed and some are significant while others are not.
Of interest will be leveraging noCorrelation along with causesNoChange in ambiguous cases, i.e., those where we have multiple relationships between two entities that are not in agreement, either by leveraging annotations to understand the conditions where the correlation is observed/not observed or by summarizing conflicting relationships into a single weighted edge.
# In systemic lupus erythematosus patients, TNFSF13B was significantly correlated with disease activity, but no significant correlation was found between IL10 and disease activity. path(DO:“systemic lupus erythematous”) noCorrelation p(HGNC:IL10) # In patients with rheumatoid arthritis, plasma leptin concentrations did not correlate with BMI. bp(MESH:“Body Mass Index”) noCorrelation a(CHEBI:Leptin)
Is an abbreviation necessary? ‘no’ would be parallel to ‘pos’ (positiveCorrelation) and ‘neg’ (negativeCorrelation).
The new language feature is: noCorrelation
As with the other BEL correlation relationships (positiveCorrelation, negativeCorrelation, association) the order of the Subject and Object terms does not matter in statements using the noCorrelation relationship.
A noCorrelation B and
B noCorrelation A are equivalent statements.
This proposed feature is for 2.0.0+ and will be added to the language so does not require any efforts to migrate older BEL knowledge.
The reference implementation must be completed before any BEP is given status “Final”, but it need not be completed before the BEP is accepted. While there is merit to the approach of reaching consensus on the specification and rationale before writing code, the principle of “rough consensus and running code” is still useful when it comes to resolving many discussions of BEL Language and API details.