Modeling Biological Networks

< Previous | Page 2 of 35 | Next >

IV.4 BACKGROUND AND SIGNIFICANCE:

Modern biology has provided a wealth of knowledge about individual cellular components and their functions. Typical experiments have carefully examined a limited number of individual components in a biological system, built hypotheses based on empirical observations, and experimented further to test these hypotheses. Researchers have deliberately restricted their analyses to well-defined systems with relatively few components, focusing on the behavior of individual molecules. While researchers have identified many functional modules or pathways, the search for modules has been haphazard.

Despite the enormous success of this approach, we can only rarely attribute a discrete biological function to an individual molecule. Indeed, most biological functions arise from complex interactions among various components (individual proteins, nucleic acids, small molecules, etc.). We need approaches that address biological complexity more comprehensively. The success of genomics in determining the entire DNA sequences for many organisms allows the definition of their gene portfolios. Extrapolation between genomes has accelerated the definition of what amounts to a "parts catalog" of cellular components in many organisms. Also, large-scale microarray and proteomics studies of the effects of systematic gene disruption and of expression levels of genes under different conditions provide data to validate the predictions of global analyses. These techniques and the emerging field of proteomics, which attempts to map all participating molecules and their interactions require techniques for interpretting and refining this enormous mass of data.

In turn, these advances have created an unprecedented opportunity to develop comprehensive explanations for biological mechanisms, in part through the identification of the fundamental constraints that limit cell behavior. While the datasets available to us are still incomplete, they suffice for analysis, model development, and prediction through model simulation. We propose an integrated program to identify these underlying constraints and to model quantitatively the structure and function (including regulatory properties) of the complex biological networks that maintain function in organisms. The emergence of databases containing integrated data on the topology of biologically significant networks and advances in understanding and quantifying the topology of complex (non-biological) networks aid this goal. The tools we will develop to study complex networks will apply to a wide range of topics of interest to the Consortium's participants, including chemotactic, neural, metabolic, signaling, and regulatory networks.