Connecting electronic health records to a biomedical knowledge graph to link clinical phenotypes and molecular endotypes in atopic dermatitis.

Precision medicine is defined by the U.S. Food & Drug Administration as "an innovative approach to tailoring disease prevention and treatment that considers differences in people's genes, environments, and lifestyles". To succeed in providing personalized medicine to patients, it will be necessary to integrate medical, biological and molecular data in order to identify all complex disease subtypes and understand their pathobiological mechanism. Since biomedical knowledge graphs (BKGs) are limited to the integration of prior knowledge data and do not integrate real-world data (RWD) that would allow for the incorporation of patient level information, we propose a first step towards using RWD, BKGs and graph machine learning (ML) to enable a fully integrated precision medicine strategy. In this study, we established a link between RWD and a BKG. Our methodology introduced a novel patient representation using graph ML applied to the BKG. This approach facilitated the interpretation and extension of ML findings, particularly in disease subtype identification with molecular data contained in the BKG. We applied our innovative methodology to deepen our understanding of atopic dermatitis, a condition with a complex underlying pathophysiological mechanism. Through our analysis, we identified seven subgroups of patients each characterized by clinical and genomic characteristics.