An ontology-based classification of Ebstein's anomaly and its implications in clinical adverse outcomes

Ebstein's anomaly (EA) is a rare congenital heart disease with significantly phenotypic heterogeneity, accompanied with multiple associated phenotypes. The classification of cases with EA based on a standardized vocabulary of phenotypic abnormalities from Human Phenotype Ontology (HPO) and its association with adverse clinical outcomes has yet to be investigated. We developed a deep phenotyping algorithm for Chinese electronic medical records (EMRs) from the Fuwai Hospital to ascertain EA cases. EA-associated phenotypes were standardized according to HPO annotation, and an unsupervised hierarchical cluster analysis was used to classify EA cases according to their phenotypic similarities. A survival analysis was conducted to study the association of the HPO-based cluster with survival or adverse clinical outcomes. The ascertained EA cases were annotated to have a single or multiple HPO terms. Three distinct clusters with different combinations of HPO term in these cases were identified. The HPO-based classification of EA cases was not significantly associated with survival or adverse clinical outcomes at a mid-term follow-up. Our study provided an important implication for studying the classification of congenital heart disease using HPO-based annotation. A long time follow-up will enable to confirm its association with adverse clinical outcomes. •Design a phenotyping algorithm of ascertainment for Ebstein's anomaly (EA) cases from Chinese electronic medical records;•Annotate phenotypes using human phenotype ontology (HPO) and classify EA patients based on phenotypic similarity;•Study the association of HPO-based cluster with the survival and clinical adverse outcomes.