karafecho
commented
3 years ago As I mentioned on the last Clinical Data Committee call, Exposures Provider is considering a new use case on 'rare disease phenotypes'. This is an idea that I had to (1) overcome the challenges that clinical KPs face when trying to use EHR data to study rare diseases and (2) better align the Exposures Provider with the goals of the Translator program (i.e., move away from common diseases such as asthma toward rare diseases).
The general idea is as follows. Rare diseases typically do not have a definitive ICD code, and EHR data on patients with rare diseases generally cannot be exposed due to privacy concerns. However, rare diseases are often associated with clusters of phenotypes that do have ICD and LOINC codes associated with them. Moreover, ontologies such as MONDO are working to map rare-disease 'concepts' or 'leaf terms' (terminal nodes in ontology trees) to ICD and LOINC codes. These EHR codes can be used as 'selection criteria' when creating a cohort of patients.
I can envision the integration of clinical data with genetics data as part of the proposed 'rare disease phenotype' use case. For instance, hypermobility-EDS is associated with phenotypes such as generalized joint hypermobility, abdominal pain, soft velvety skin, and bruising. A causal genetic variant(s) for hEDS has not been established. The Genetics Provider may be able to suggest variants that are shared by hEDS-associated phenotypes.
@flannick : Thoughts?
This issue is intended to initiate discussion on how Translator clinical KPs can best leverage the genetics knowledge available via the Translator Genetics Provider.