High-throughput expression profiling enables the global study of gene activities. Genes with positively correlated expression profiles are likely to encode functionally related proteins. However, all biological processes are interlocked, and each protein may play multiple cellular roles. Thus the coexpression of any two functionally related genes may depend on the constantly varying, yet often-unknown cellular state. To initiate a systematic study on this issue, a theory of coexpression dynamics is presented. This theory is used to rationalize a strategy of conducting a genome-wide search for the most critical cellular players that may affect the coexpression pattern of any two genes. In one example, using a yeast data set, our method reveals how the enzymes associated with the urea cycle are expressed to ensure proper mass flow of the involved metabolites. The correlation between ARG2 and CAR2 is found to change from positive to negative as the expression level of CPA2 increases. This delicate interplay in correlation signifies a remarkable control on the influx and efflux of ornithine and reflects well the intrinsic cellular demand for arginine. In addition to the urea cycle, our examples include SCH9 and CYR1 (both implicated in a recent longevity study), cytochrome c1 (mitochondrial electron transport), calmodulin (main calcium-binding protein), PFK1 and PFK2 (glycolysis), and two genes, ECM1 and YNL101W, the functions of which are newly revealed. The complexity in computation is eased by a new result from mathematical statistics.gene expression ͉ microarray ͉ urea cycle ͉ correlation ͉ glycolysis M icroarrays have generated an enormous amount of geneexpression data from a variety of biological studies (1-5). After proper preprocessing, the data can be stored as a matrix of real numbers with N rows and m columns. Rows represent the gene-expression profiles, and columns represent the cell types, time points, or environmental or other experimental conditions under which the mRNA samples are taken. To elucidate microarray data, most methods (6-10) rely on the notion of profile similarity as described for Fig. 1. It is argued that coexpressed genes are likely to encode proteins that participate in a common structural complex, metabolic pathway, or biological process (6, 10).Despite the many successful applications reported in the literature, there is an important issue that is hard to address by profile-similarity analysis. As is known, all biological processes are interlocked, and many proteins have multiple cellular roles. Two proteins engaged in a common process under certain conditions may disengage and embark on activities of their own under other conditions, which implies that both the strength and pattern of association between two gene profiles may vary as the intrinsic cellular-state changes. Weaver (11) discussed two transcription factors, Max and thyroid hormone receptor (TR). They can serve either as activators or repressors depending on other molecules bound to them. Max can bind to Myc and form a M...