Content area
Abstract
A general framework is presented for using genome- and proteome-wide measurements to build, assess, and refine a model of a gene-regulatory network. This framework involves (i) systematically perturbing the known genetic and environmental components of the network; (ii) observing the global gene-expression response to each perturbation; (iii) inferring hypothetical network models that are consistent with the observed responses, optionally guided by a preexisting model or other prior knowledge; (iv) designing new perturbations able to distinguish between alternate models and repeating steps (i–iv), thereby refining the network model over successive iterations of perturbation and global measurement. This framework was initially explored in simulation, by analyzing random networks with varying numbers of genes N and interactions per gene k. Although the number of inferred networks increased exponentially with N, the number of inferred networks decreased exponentially with the number of additional perturbations performed. Motivated by these theoretical results, this framework was used to analyze 20 systematic perturbations to the yeast galactose-utilization pathway. Global changes in mRNA and protein expression were measured in response to each perturbation using whole-genome DNA microarrays and mass spectrometry; novel statistics, based on maximum-likelihood estimation, were used to determine which changes were significant. These data were compared to expression changes as predicted by a qualitative model of galactose-utilization, and possible model refinements suggested by these comparisons were evaluated through additional rounds of perturbation and global measurement. Finally, to address observed interactions between galactose utilization and other metabolic pathways, the basic model was extended using existing libraries of protein-protein and protein-DNA interactions. This framework and associated proof-of-principle in yeast provides a road map for how systems biology will be applied to more complex biological systems.
