Abstract
Non-alcoholic fatty liver disease (NAFLD) is a spectrum of diseases ranging from simple steatosis without inflammation or fibrosis to nonalcoholic steatohepatitis (NASH). Despite the strong association between dietary factors and NAFLD, no dietary animal model of NAFLD fully recapitulates the complex metabolic and histological phenotype of the disease, although recent models show promise. Although animal models have significantly contributed to our understanding of human diseases, they have been less successful in accurate translation to predict effective treatment strategies. We discuss strategies to overcome this challenge, in particular the adoption of big data approaches combining clinical phenotype, genomic heterogeneity, transcriptomics, and metabolomics changes to identify the ideal NAFLD animal model for a given scientific question or to test a given drug. We conclude by noting that novel big data approaches may help to bridge the translational gap for selecting dietary models of NAFLD.