Fig. 4. Responses to clinical review questions for disease-phenotype associations.

These are for the 10 diseases marked by the clinical reviewers, separated by their membership in SMP and BDLP. If an association was recovered by both databases, it is included in both sets here. Reviewers were blinded to the source of the phenotype association. a Proportion of clinician responses to the question “Is this association established in literature, treatment guidelines, or policy discussing this disease?” across the set of ten disease responses. b Proportion of clinician responses to the question “What kind of association is this?” across the set of ten disease responses. c Proportion of clinician responses to the question “How often do you recognise this association in the course of your clinical practice for this disease?” across the set of ten disease responses.