Z Gastroenterol 2016; 54 - KV185
DOI: 10.1055/s-0036-1586961

Prioritizing drug targets for non-alcoholic fatty liver disease based on comorbidity network analysis

TP Nguyen 1, T Sauter 1, L Caberlotto 2, F Lammert 3, JG Schneider 1, 3
  • 1Université du Luxembourg, Luxembourg, Luxembourg
  • 2The Microsoft Research, University of Trento Centre for Computational Systems Biology, Rovereto, Italy
  • 3Saarland University Medical Center, Department of Medicinne II, Homburg, Deutschland

Background and aims: Non alcoholic fatty liver disease (NAFLD) is highly associated with other components of the metabolic syndrome, in particular obesity and diabetes. However, systematic network medicine approach of the complex phenotypic dependencies and comorbidities has yet to be performed. Our aim now is to identify novel drug targets for NAFLD treatment by a systematic in silico network analysis.

Methods: We constructed an interaction network of curated genes relevant to NAFLD and its comorbid diseases of interest. Network mining was applied to characterize the NAFLD comorbidity network, and identify the key genes based on centrality ranking. Large-scale shortest path analyses showed NAFLD connectivity to other comorbid diseases and allowed inferring the molecular mechanisms underlying comorbidities.

Results: Overall, we identified 594 disease genes, most of which (38.7%) were related to glucose metabolism disorders (n = 230). Lipid metabolism disorders were represented by 116 genes, and obesity by 114 genes; 19 genes were associated with metabolic syndrome, and 30 genes with non-alcoholic fatty liver disease (NAFLD). Network construction on the 594 disease genes resulted in a network of 2,175 proteins and 4,605 interactions from the human protein reference database. Among the most central genes in the network, 2 genes were directly related to NAFLD (TGFB1 and B2) and 8 new candidate genes were inferred. We identified 10 key genes that were related to at least two diseases and then consecutively regulated the comorbidities between the diseases. Of note, most of the 10 genes were related to lipid and glucose metabolism, and two (INS, IRS1) were involved in aldosterone-regulated sodium absorption and diabetes. Furthermore, we found a large number of paths linking the NAFLD-related genes to glucose and lipid metabolism disorders and obesity (46,143, 22,000 and 21,516 paths, respectively), in particular genes related to adipocytokine and PPAR signaling and cytokine-cytokine receptor interaction.

Conclusions: The comorbidity network analysis of NAFLD allowed the identification of genes that could represent the most promising molecular targets for prioritization of drug therapy in NAFLD.