Fragile dynamics enable diverse genomic determinants to influence arrhythmia propensity

Genotype-phenotype relationships are determinants of human diseases. Often, we know little about why so many genes are involved in complex common diseases. We hypothesized that this multigene effect arises from the relationship between genes and physiological dynamics. We tested this hypothesis for arrhythmias as physiological dynamics define this disease. We integrated graph theory analysis of genomic and protein-protein interaction networks with dynamical models of ion channel function to identify the physiological dynamics of genome wide variation for five different arrhythmias. Regulatory networks for the cardiac conduction system and arrhythmias were constructed from GWAS and known disease genes. Electrophysiological models of myocyte action potentials were used to conduct extensive parameter variations to identify robust and fragile kinetic parameters that were then, using regulatory networks, associated with genomic determinants. We find that genome-wide determinants of arrhythmias that represent many cellular processes are selectively associated with fragile physiological dynamics of ion channel kinetics. This association predicts disease propensity. Deep RNA sequencing from human left ventricular tissue of arrhythmia and control subjects confirmed the predictive relationship. Taken together these studies show that the varied multigene effects of arrhythmias arises because of associations with fragile kinetic parameters of cardiac electrophysiology.\n\nSignificance StatementOur understanding of the genetics of common diseases has advanced exponentially over the past decade. We now know that differences and variation in multiple genes contribute to disease susceptibility with significant heterogeneity in the phenotype. However, how genetic variation contributes to disease phenotypes remains unknown. We hypothesized that the relationships between physiological dynamics and genetic architecture is a fundamental determinant of disease susceptibility and genetic heterogeneity. To test our hypothesis, we integrated mathematical models of cardiac electrophysiology with genetic network models of cardiac arrhythmias. We found that disease related genome variants were selectively associated with fragile kinetic parameters that predict disease propensity and identified several novel cellular processes associated with arrhythmogenesis.