(475g) Qualitative Multiscale Methodology for Biological Networks Driven by Experimental and Clinical Data: Insulin Network in Beta Cells and Association with Diabetes

There is a significant increase in genomics, metabolomics and proteomics datasets and a growing demand to provide qualitative mathematical models for fighting obesity, diabetes and other diseases related to the metabolism and regulation. Qualitative approaches are efficient and adequate in many cases where there is limited quantitative information on reaction mechanisms for biological networks.

We have designed a multiscale methodology using network ideas (molecules-cells-tissues-body) to model biological networks in different length and time scales extending the simple boolean networks and avoiding complex stoichiometric approaches. Our method allows the creation and customization of networks based on limited user data and available bioinformatics databases. The discrete time steps, molecular states, and relative concentrations are described like a cellular automata model where each node(the molecule state and properties) is updated from the input and signals of the neighboring connected ones. The connections between the molecular nodes can be represented in an experimentally realistic way using linear or nonlinear connections allowing the description of complex mechanisms such as feedback loops.

We focus on the role of insulin which is made and released from the beta cells and controls the level of glucose in the blood. Insulin is associated with both types of diabetes and there are currently a lot of available data and a need to explain the complex interconnected mechanisms causing the disease. Simulations of signaling and metabolic pathways of insulin in beta cells and pancreas tissue will be presented and a comparison with experimental data and other approaches will be discussed.

BIBLIOGRAPHY: System Modeling in Cellular Biology: From Concepts to Nuts and Bolts, Edited by Zoltan Szallasi, Jörg Stelling and Vipul Periwal, 2006