Systems Analysis and Translational Research in Metabolism

The term “systems analysis” in biology has had multiple definitions.  Recently it has referred to the use of great computing power to ascertain and analyze immense data sets, generally related to the genome, the proteome and/or the metabolome.  By this approach it is generally assumed that the availability of massive data and computing power will lead to greater understanding of metabolic regulation; the approach is sometime referred to as “hypothesis-generating.”  But, the term systems analysis had an earlier definition – the expression of physiological knowledge in mathematical models which in themselves represent specific hypotheses which can be carefully tested and revised to yield a comprehensive representation of physiological understanding.   This earlier approach has much in common with the traditional scientific method in which hypotheses and experiments are synergized to yield important and usable physiological understanding.   We used careful measurements made in animals to examine the relationship among plasma measurements during physiological perturbations.  We imposed the concept of optimal simplicity – a mathematical representation which could account for the data with the simplest physiological representation.  The model, of course, explained the data used for its construction; more important it made specific predictions which were amenable for experimental test.   From the model were able to define a series of important clinical variables: insulin resistance and functionality of the beta-cells of the endocrine pancreas.  A single parameter emanating from the model – the “disposition index” became very useful because it was shown to be the strongest predictor of conversion from pre-diabetes to full blown Type 2 diabetes mellitus.  It was also apparently coded by several genetic variants.  The model was recently modified to include the role of the liver to phosphorylate plasma glucose to lactate – this modification allows us to examine the role of insulin-independent glucose utilization in the pathogenesis of Type 2 diabetes.  Other outcomes of the model include understanding of insulin transport across capillary endothelium which can be altered in the obese, insulin resistant state.  These studies of carbohydrate metabolism are a case study of hypothesis-based systems analysis, and demonstrate that the more traditional definition of “systems analysis,” – hypothesis based, rather than hypothesis-generating remains a powerful approach to understanding physiological regulation, and how it can be used to advance clinical medicine.