Figure 2: Leveraging Data Analysis and Integration to Elucidate AF Mechanisms and Classify Patients.

Genetic, transcriptomic and other -omic studies generate enormous quantities of data on AF, as well as important clinical characteristics and endophenotypes. Using machine learning approaches, bioinformatics and polygenic risk scores will help to create new phenotypic classes and stratify patients based on molecular mechanisms. These groupings can then serve as inputs to subsequent rounds of analysis in an iterative fashion to elucidate increasingly refined mechanistic underpinnings of AF, which may, in turn, lead to personalised medicine with different treatment modalities for specific subgroups of patients. PRS = polygenic risk scores.