(A) Module #126 is defined by data from 30 different studies and contains a highly coherent set of 11 genes, all but two of which are known to be involved in late mitosis and in cell septation. (B) Regulation by coactivators and corepressors. We plot the average module expression (red) and the background genome-wide mean and standard deviation for a random set containing 11 genes (green) under conditions in which the Swi/Snf complex are not expressed in minimal and rich media (Sudarsanam et al, 2000), in conditions blocking components of the RSC complex (Angus-Hill et al, 2001) and in a strain lacking Bur6, a component of the NC2 cofactor (Geisberg et al, 2001). There is significant induction in all Swi/Snf conditions and in the NC2 experiment, indicating a possible negative role for these cofactors in regulating the module. There is also a significant repression in the rsc3 strain (but not in the rsc30 strain), indicating that RSC may have a positive role in the regulation of the module. (C) Extended regulatory model for the cytokinesis module. See text for details.

(A) Outline of the hyperosmotic stress signaling pathway. Two Hog-dependent (Ssk1, Ste11) and one Hog-independent (Msn2/4) pathways mediate the hyperosmotic stress signal. (B) Response of selected modules to osmotic stress. We plot the average expression of several modules that our algorithm associated with osmotic stress conditions, in several strains knocked out for key players in the HOG pathway. The graphs show modules' mean expression time courses after treatment with 0.5 M KCl. In general, modules #232 (Ribosomal proteins) and #524 (RNA processing), #686 (Amino-acid biosynthesis), #503 (Purines) and #985 (Ergosterol biosynthesis) are repressed as part of the ESR, with peak response observed at 20 min and re-establishment of normal transcription after 40–60 min. Modules #536 (Respiration) and #1215 (Gluconeogenesis) are induced with similar kinetics. Specific modules show particular deviation from these two general trends. (C) Multiple signals additively regulate module #524. We plot the mean expression of module #524 and its standard deviations in four strains (wt, hog1, ste11, ssk1) under two levels of hyperosmotic shock (0.5 and 0.125 M KCl). There is marked difference between the ssk1 and hog1 strains and the wt, ste11 strains, suggesting the existence of two regulatory mechanisms. An osmotic stress-specific, Ssk1/Hog1-mediated signal represses the module in both low and high levels of osmotic shock. In high osmotic shock, a second, Hog1-independent signal (which is probably related to the general ESR) is active in parallel to the Hog1 signal and contributes additively to the repression of the module. (D) A two-phase regulatory program for module #536. We show the time courses of the mean expression of module #536 (Respiration) and its main regulator Hap4, when treated with 0.5 M KCl in the wt strain. The module exhibits weak and poorly correlated induction, which is Hap4 independent, during the primary phase of the osmoregulation program (0–40 min). A second phase is observed at 60–180 min, where a tightly correlated induction is facilitated by increase in HAP4 expression.

(A) Bicluster analysis. Our SAMBA biclustering algorithm analyzes an integrated data set to discover an extensive collection of modules. Each module consists of a set of genes and is supported by a set of functional properties. We analyze novel data sets by testing their effect on the modules derived from the entire compendium. (B) Modules' dimensions. The distribution of the number of genes and properties in each module indicates that modules are characterized by specific sets of genes (10–50) and a large number of different experiments (20–100). (C) Synergism among studies. The graph shows the distribution of the number of studies contributing to each module. A total of 86% of the modules use data from more than one study. (D) The module–condition view. To obtain a global view of the behavior of our modules across all conditions, we clustered the module mean values across all conditions. Rows represent modules and columns represent conditions, with numbered colored bars indicating the study reporting each condition (the full references of the studies are available on the website). We show low means in green and high means in red. The global view reveals that the massive repression and induction of genes in stressful conditions dominates the compendium. Using our integrative analysis, we can dissect this response into components and study their specific regulation. The numbers on the right refer to modules addressed in this study.

(A) Response of selected modules to disruptions in the GAL system. We plot the mean and standard deviation of the expression of several key modules that our algorithm associated with conditions from the galactose data set (Ideker et al, 2001). For each module, we plot the behavior in four galactose-related mutants and two double mutants grown with (red) and without (green) galactose. Module #389 (Galactose metabolism) is strongly repressed when galactose is lacking or when the GAL pathway yield is compromised. Modules #232 (Ribosomal proteins) and #524 (RNA processing) are repressed when growth is slower. Interestingly, module #524 is particularly repressed when gal4 is knocked out, and is induced when gal80 is knocked out and galactose is not available (right-most bars). Modules #536 (Respiration) and #1215 (Gluconeogenesis) are induced when galactose is not available or not processed. Here again, the gal80 mutants exhibit altered behavior (module #536 is repressed). Modules #503 (Purines), #686 (Amino-acid biosynthesis) and #967 (Methionine metabolism) are repressed when growth is slower. (B) Disrupted coupling of two stress-related modules. The plot shows the mean expression of modules #232 (Ribosomal proteins) and #524 (RNA processing) in the galactose pathway experiments, together with the linear regression line of the dependency between the mean expression levels of the two modules over the entire compendium (Materials and methods). Broken lines indicate ±1 standard deviation. The gal80 mutants exhibit increased expression, while the gal4 mutants (excluding gal80gal4) exhibit decreased expression relative to the compendium trend, supporting a possible involvement of the gal80–gal4 circuit in the regulation of the modules. (C) Hap4-independent repression of module #536 in gal80 strains. We plot the mean expression of module #536 (Respiration) and the expression of the gene coding for its direct regulator Hap4 in selected conditions. When galactose is not available, module #536 is induced via increased expression of HAP4. Similar effect is observed in several other conditions, for example in the gal4 strain. In gal80 strains, we observe repression (or lack of induction) of the module, although the HAP4 gene is expressed at high levels.

According to the current prevalent paradigm (top part), novel data are analyzed in isolation, typically using clustering and expert manual analysis of specific clusters. We suggest a new approach (lower part) in which the community maintains the current publicly available data sets and the set of biological modules revealed by them. Modules may cover all aspects of biological processes and their regulation, as revealed, for example, by our biclustering algorithm. Using this resource, novel data sets can be represented in terms of the behavior of known and novel modules, providing an objective and transparent method for understanding, communicating and reusing high-throughput data.