ABSTRACT
When realistic mathematical models of whole body metabolism eventually become available, they are likely to add entirely new dimensions to the understanding of the integrated physiological function of the organism, in particular the mechanisms governing the regulation of transitions between different physiological states, like fed-fasted, exercise-rest and normal-diseased. So far the strategy for whole body modelling has primarily been a bottom-up approach where the central problem is an apparently insurmountable barrier of complexity involved in defining and optimising the huge number of parameters. Here we follow a top-down strategy and present a complete mathematical framework for realistic whole body model development. The approach proposed is modular and hierarchical and whole body metabolism is taken as the top level. Next are the organs, where the sum of the contributions from the individual organs must equal the top level metabolism. This hierarchy can be extended to lower levels of organisation, i.e. clusters of cells, individual cells, organelle and individual pathways. Exploiting this hierarchy, metabolism at each level forms an absolute constraint on the contributions from lower level. Importantly, these constraints can in many ways be defined experimentally through mass balance and flux data. Furthermore, the constrained approach allows the lower level models to be developed independently and subsequently adapted to the whole body model. The paper describes the process of whole body modelling in practical terms, centred on a mathematical framework, devised to allow whole-body models of any complexity to be developed. Furthermore, an example of sub-model incorporation in the whole-body framework is illustrated by adapting an existing erythrocyte model to the whole body constraints. Finally, we illustrate the operation of the system by including two sets of whole-body data from humans, reflecting two different physiological states.