- Open Access
Design principles of molecular networks revealed by global comparisons and composite motifs
© Yu et al.; licensee BioMed Central Ltd. 2006
- Received: 16 March 2006
- Accepted: 20 June 2006
- Published: 19 July 2006
Molecular networks are of current interest, particularly with the publication of many large-scale datasets. Previous analyses have focused on topologic structures of individual networks.
Here, we present a global comparison of four basic molecular networks: regulatory, co-expression, interaction, and metabolic. In terms of overall topologic correlation - whether nearby proteins in one network are close in another - we find that the four are quite similar. However, focusing on the occurrence of local features, we introduce the concept of composite hubs, namely hubs shared by more than one network. We find that the three 'action' networks (metabolic, co-expression, and interaction) share the same scaffolding of hubs, whereas the regulatory network uses distinctly different regulator hubs. Finally, we examine the inter-relationship between the regulatory network and the three action networks, focusing on three composite motifs - triangles, trusses, and bridges - involving different degrees of regulation of gene pairs. Our analysis shows that interaction and co-expression networks have short-range relationships, with directly interacting and co-expressed proteins sharing regulators. However, the metabolic network contains many long-distance relationships: far-away enzymes in a pathway often have time-delayed expression relationships, which are well coordinated by bridges connecting their regulators.
We demonstrate how basic molecular networks are distinct yet connected and well coordinated. Many of our conclusions can be mapped onto structured social networks, providing intuitive comparisons. In particular, the long-distance regulation in metabolic networks agrees with its counterpart in social networks (namely, assembly lines). Conversely, the segregation of regulator hubs from other hubs diverges from social intuitions (as managers often are centers of interactions).
- Regulatory Network
- Metabolic Network
- Assembly Line
- Action Network
- Additional Data File
Importance of the four networks
We chose these four networks because they are the most commonly studied networks in yeast and because they can be easily related to the central dogma of molecular biology, which describes the basic (genetic) information flow in a cell. There are also other types of biological networks, such as synthetic lethal networks and chromosomal order networks [10, 11]; however, these networks do not overlap with the central dogma and are, therefore, not the focus of this paper. Furthermore, most of these networks are not suitable for large-scale topological analysis because we do not have enough information on them.
Another important reason for us to choose these four networks is that there are many appealing analogies between these biological networks and corresponding social networks [12–14]. Because people have clear intuition for social networks, based on daily experiences, these analogies can make molecular networks easier to comprehend. For example, social hierarchy networks resemble the regulatory networks in that they define who has to obey orders from whom. Social acquaintance networks describe who is known to whom in the society and are, therefore, similar to interaction networks in biology [13, 14]. Finally, enzymes at different steps of the metabolic network can be considered as workers at different steps of the assembly line in a factory.
Composite features in combined networks
Individual networks have been globally characterized by a variety of graph-theoretic statistics (Additional data file 1), such as degree distribution, clustering coefficient (C), characteristic path length (L) and diameter (D) [12, 15, 16]. Barabási and Albert  proposed a 'scale-free' model in which most of the nodes have very few links, with only a few of them (hubs) being highly connected. In addition to topological statistics and hubs, network motifs provide another important summary of networks. These are over-represented sub-graph patterns in networks, and they are considered as basic building blocks of large-scale network structures . Recently, Yeger-Lotem et al.  combined the interaction and regulatory networks in yeast and searched for patterns in the combined network.
Here, we build on previous network studies and extend them in novel directions by combining all four networks in our analysis. Our goal is to examine the topological features of our combined network. We call these 'composite features' to distinguish them from those in single networks (see Materials and methods). By analyzing these in all four networks, we were able to find some basic principles characterizing biological networks. For example, previous studies have shown most biological networks are scale-free, having only a few hubs as the most important and vulnerable points [12, 15]. It is quite reasonable to assume that our four networks will share the same set of hubs as explained in detail below. However, we analyzed the composite hubs among the four networks and showed that the regulatory network tends to use a distinctly different set of hubs compared to the other three networks. Furthermore, one fundamental question in biology is how the cell uses transcription factors (TFs) to regulate and coordinate the expression of thousands of genes in response to internal and external stimuli [8, 19–21]. Through examining composite motifs, we could potentially shed some light on this question. In particular, we show that the expression of enzymes at different steps of the same pathway tends to have time-delayed relationships mediated by inter-regulating TFs.
Overall comparisons of all four networks
We calculated many topological statistics in all four networks, which are summarized in Figure 1a. All four networks display 'scale-free' and 'small-world' properties. However, the regulatory network is different from other networks in that its clustering coefficient is exceptionally small. This is because most of the target genes are not TFs. Therefore, the target genes of the same regulator tend not to inter-regulate one another. Moreover, since the regulatory network is directed, it is divided into regulator and target sub-networks when calculating the degree distribution. It has been shown that the regulator network is a scale-free network. But, the target network might have an exponential degree distribution, instead . This means that there are no hubs in the target network. Therefore, when we examined the hubs and composite hubs in the regulatory network, we focused only on the regulator population. This also makes sense biologically, because we are more interested in how a gene's expression is regulated in different networks; the regulators (that is, TFs) are the ones that carry out the regulatory functions.
Furthermore, we analyzed the relationships between different networks. Since the relative position of nodes in a network is one of the most important features of the network, we examined the relationships between networks using their distance matrices, that is, distances between all protein pairs. We divided all pairs of proteins in a network into three groups: connected pairs; close pairs (distance = 2); and distant pairs (distance ≥3). We used Cramer's V, a measurement derived from χ2 statistics, to examine the association between networks, that is, whether pairs of proteins in one group of a network tend to be in the same group of another network. Our calculations confirm that all networks are indeed significantly related to each other (Figure 1b). We also tried many other metrics of relatedness - for example, Pearson correlation coefficient, mutual information, contingency coefficient, and association score. They all show similar results (see Supplementary Table 1 in Additional data file 1).
Composite hubs tend to be more essential than hubs in single networks
Scaffolding of the regulatory network is different from other networks
Because all four biological networks are scale-free (Figure 1a; here we only consider the regulator population within the regulatory network), it can be shown that they should share the same hubs by chance alone due to hubs' essentiality (Additional data file 1). It is interesting to see whether this is indeed the case for biological networks, that is, whether they are built on the same scaffolding.
Our calculation shows that the scaffolding of three networks (metabolic, interaction and co-expression) tends to be the same, that is, hubs in one network tend to overlap with those in another when compared to random expectation (Figure 3b). The results agree with previous studies showing that interacting proteins tend to be co-expressed [27–30]. Furthermore, we calculated the random expectation by taking into consideration the fact that hubs tend to be essential [15, 24]. We found that the hub overlap between networks could not be explained by simply considering the essentiality of hubs (Supplementary Figure 2 in Additional data file 1).
Surprisingly, hubs in the regulator network do not have the tendency to be hubs in other networks. Though counter-intuitive, this observation is reasonable in that most TFs and their targets do not tend to be co-expressed , and most TFs are unlikely to interact with their targets. Therefore, we divided the four networks into two classes: regulation and action. The action networks include the interaction, co-expression and metabolic networks. It is clear that the cell separates the regulatory network from the action networks. Since all action networks are governed by the regulatory network as discussed below, the separation potentially could provide stability to the cell (Supplementary Figure 5 in Additional data file 1).
Here we have excluded the comparison between regulator and metabolic networks because the two networks only share one common protein. It is possible to argue that our definition of hubs is somewhat arbitrary. But all results remain the same even when we used different cutoffs to define hubs. We further tested the functional composition of the overlapping proteins among networks, which is similar to that of each individual network and random expectation (Supplementary Figures 3 and 4 in Additional data file 1).
Neighboring pairs in all action networks are co-regulated
Above, we separated the regulatory network from the others; now we show that the three action networks can be further subdivided into two groups (that is, short-range and long-range) based on how the genes in them are regulated by TFs. We investigated this through looking at composite motifs within the combined regulatory-action network. We focused on a few key motifs, which we call triangles, trusses, and bridges (see Materials and methods).
In a truss, two proteins share the same feed-forward loop (FFL; Figure 4b). FFLs are robust against noise . Previous work has also shown that genes co-regulated by more than one regulator tend to be tightly co-expressed . Therefore, trusses are designed to maintain stable co-expression between gene pairs. Their biological function is similar to that of triangles.
We examined the distributions of the enrichment of trusses in all three combined networks. As expected, the three distributions share similar patterns with that of triangles (Figures 4a,b). In all distributions, only connected pairs show enrichment of trusses, which further confirms the biological function of trusses. Given the fact that the regulatory network in yeast is far from complete, we believe that many actual trusses are missed by our analysis because some of the edges are missing in our dataset. To confirm this, we also looked at semi-trusses. A semi-truss is a truss with only one FFL (Figure 4c). We believe that many of these semi-trusses are actually full trusses given the incomplete nature of our dataset. Figure 4c shows highly similar results to those in Figure 4b, thus providing support for our conclusion.
Interestingly, it has been shown experimentally that triangles and trusses can also generate temporal programs of expression by having serial activation coefficients with different targets, which is quite intuitive and reasonable [33, 34]. It should also be noted that some FFLs ('incoherent FFLs') could provide pulses and speeding responses, although the majority of FFLs are coherent, acting as 'persistence detectors' [35, 36].
Distant enzymes in the same pathway tend to have delayed expressions mediated by regulator bridges
We then analyzed the composite motifs in the combined metabolism-co-expression network. Figure 5b shows that co-enzymes tend to be co-expressed, and the tendency of co-expression decreases as the distance between the enzymes increases. On the other hand, enzymes in different steps of the same pathway tend to have expression relationships other than co-expression, typically time-delayed relationships (Supplementary Figure 7c in Additional data file 1). This tendency increases as the distance increases. The likelihood for far-away enzymes in the same pathway to have other expression relationships is significantly higher than random expectation. This observation shows that enzymes in the same pathway are not necessarily co-expressed; nevertheless, their expression needs to be well-coordinated for the whole pathway to function normally. This is the reason why bridges are enriched in disconnected enzyme pairs in the metabolic network (Figure 5a). Similar results were also found in other time-course expression experiments , but not in the interaction network (Additional data file 1). This conclusion is further supported by a specific case study in Escherichia coli amino acid biosynthesis pathways . As we mentioned above, metabolic pathways in the cell are very similar to assembly lines in a factory. It is reasonable to assume that, without decreasing the efficiency of the whole assembly line, workers at downstream steps of the line do not have to show up for work until those at upstream steps have finished their job. Similarly, in terms of metabolic pathways, we observed that enzymes at downstream steps tend to be expressed after those at earlier steps. The bridge motifs are designed to manage such expression relationships between enzymes, and, therefore, to maintain normally functioning metabolic pathways in the cell.
Here we examine the four most commonly studied networks in yeast. Previous work has shown that social networks share common characteristics with biological networks [12–14]. Our results further confirm this. In particular, many common social networks are related. We also found that biological networks, even though seemingly quite different, are clearly related to each other. In social networks, people under the same supervisor normally know each other, and, as such, may be said to be connected in acquaintance networks. Accordingly, in the biological networks, we observed that connected pairs in action networks tend to be co-regulated. More interestingly, distant enzymes in the same pathway show a surprising tendency to have delayed expression coordinated by regulator bridges. Although this phenomenon is readily understandable through an analogy to assembly lines, it is still striking to see it so strongly manifest in real biological networks. However, the structure of biological networks obviously has some differences from that of social networks. In a normal social context, it is reasonable to assume that a supervisor knows his or her staff. Therefore, supervisors with large staffs (that is, hubs in the social hierarchy) tend to be hubs in acquaintance networks. This is not the case for biological networks: the regulatory network uses a different set of hubs than the action networks.
Recently, Mazurie et al.  also analyzed the composite network motifs in the combined regulatory and interaction network. They used a similar approach to Yeger-Lotem et al.  and examined the composite motifs that are over-represented in a strictly mathematical sense. However, they found that the overabundance of these network motifs "does not have any immediate functional or evolutionary counterpart" . These findings confirm that we should not only look at the most mathematically over-represented motifs, but that we should also focus on key, obviously functionally relevant ones, further highlighting the importance of our approach. In our analysis, we first identified composite motifs that could potentially have biological functions and examined the enrichment of these motifs in the combined network. Our results have clearly shown that the enrichment of some composite motifs is closely related with their function. For example, bridges are only enriched between far-away enzymes in the same pathway because the expression of these enzymes needs to be well coordinated.
The regulatory network was created by combining five different datasets [8, 9, 22, 31, 39, 40]. A link in the network is defined as a TF-target pair. We excluded DNA-binding enzymes (for example, PolIII) and general TFs (for example, TATA-box-binding protein) from the regulatory network.
The co-expression network was created using the microarray dataset of Cho et al. . A link here is defined as a co-expressed gene pair with a correlation coefficient larger than or equal to 0.8. It is possible to argue that the cutoff (0.8) here is somewhat arbitrary. We repeated all relevant calculations using different cutoffs ranging from 0.5 to 0.9. All results remained the same (Additional data file 1).
The interaction network was created by combining various databases and large-scale experiments [2–5, 41–43]. Because large-scale experiments are known to be error-prone , we only considered high-confidence protein pairs as true interacting pairs (likelihood ratios ≥300, P value < 10-200 as estimated by the hypergeometric distribution; likelihood ratios measure the enrichment of interacting protein pairs with certain genomic features ; see Additional data file 1 for a detailed discussion).
The metabolic network was downloaded from the KEGG database . However, the metabolic network is different from the other networks in that the nodes in the network are small molecules and they are connected by the enzymatic steps between them. To compare the metabolic network to others, we transformed the network in the following way: each enzyme was considered a node in the network, and enzymes working on adjacent steps were considered 'connected'. Whenever there is more than one enzyme in the same enzymatic step (that is, co-enzymes), we also consider all co-enzymes as 'connected'. Only main substrates and products were used to perform the transformation. Most co-factors and carriers (for example, ATP and H2O) were removed from all reactions.
All four networks are available through our supplementary website .
Composite topological features
Yeger-Lotem et al.  defined composite motifs operationally as over-represented patterns in the combined network as compared to a randomized control. Using this criterion, they exhaustively searched through the combined network and were able to detect 1 two-node, 5 three-node and 63 four-node composite motifs. A similar study has also been performed by Zhang et al. . Instead of automated detection of new composite motifs, we manually selected five basic composite motifs for further analysis because, as discussed below, these composite motifs summarize the most basic biological relationships between protein pairs within the four networks. Our analysis covered all four biological networks. We analyzed not only nearest neighbors, but also protein pairs that are further apart in each network. Most importantly, we were able to gain significant insights into the biological functions of the five composite motifs by comparing their patterns of occurrence in the combined networks.
Definition of five composite motifs
We first examined the regulatory relationships between protein pairs in action networks and created three combined networks by combining the regulatory network with each of the other three networks. We defined three biologically meaningful composite motifs in all three combined networks, based on the fact that co-regulation (that is, that two proteins share the same regulator) and inter-regulation (that is, that the regulator of one protein regulates the regulator of another protein) are the two most basic regulatory relationships between a pair of proteins. The three basic composite motifs that we defined are: co-regulation motifs (triangles); integrated FFLs (trusses); and bridging motifs (bridges) (Supplementary Figure 6 in Additional data file 1). Yeger-Letem et al.  determined that triangles and trusses are significantly overrepresented motifs, but bridges are not. However, we are able to show the biological importance of bridges in the main discussion (see above).
We also created another combined network by combining the co-expression and metabolic networks. Qian et al.  developed a local clustering method to detect four expression relationships between gene pairs: co-expressed, time-shifted, inverted, and inverted time-shifted. Using the local clustering method, we defined two composite motifs in this combined network (Supplementary Figure 7 in Additional data file 1): the co-expression motif, a pair of enzymes at distance k in the metabolic network that are co-expressed; and the shifted motif, a pair of enzymes at distance k in the metabolic network that have expression relationships other than co-expression. Most of these pairs have time-shifted relationships.
For each of the above composite motifs, we determined its degree of enrichment at different distances in different action networks in the following way. We first counted the number of protein pairs at a certain distance k in each of the three action networks. Then, we calculated the fraction of pairs that are within a certain composite motif.
Calculations of the random expectation of hub overlaps
To calculate random expectation of hub overlaps, we first created randomized networks for each biological network by randomly shuffling node degrees among proteins throughout the whole network. In this manner, the degree distributions of the original networks are conserved in randomized networks. Then, we calculated the overlap of hubs between the randomized networks of the two original networks. The procedure was repeated 1,000 times. The average overlap is considered as the random expectation.
An observed enrichment in hub overlap can be partly explained by the fact that hubs tend to be essential. In order to take into consideration hub essentiality, we created randomized networks by shuffling degrees only among genes that are either essential or non-essential. In this manner, the tendency for hubs to be essential is conserved in randomized networks. Other steps are the same as above.
Similarly, an observed enrichment in essentiality of composite-hubs compared to hubs in a single network can be at least partly explained by the fact that hubs generally tend to be essential. To prove this, we again created randomized networks where the tendency for hubs to be essential is conserved. We then compared observed essentiality enrichment in composite-hubs with calculations based on the randomized networks.
The following additional data are available with the online version of this paper. Additional data file 1 is a PDF file containing the supplementary materials to the main manuscript, in which we introduce the details of many calculations performed in the main text and discuss many additional results supporting the conclusions in the main text.
This work is supported by a grant from NIH/NIGMS (P50 GM62413-01).
- Hartwell LH, Hopfield JJ, Leibler S, Murray AW: From molecular to modular cell biology. Nature. 1999, 402 (6761 Suppl): C47-52. 10.1038/35011540.PubMedView ArticleGoogle Scholar
- Ito T, Tashiro K, Muta S, Ozawa R, Chiba T, Nishizawa M, Yamamoto K, Kuhara S, Sakaki Y: Toward a protein-protein interaction map of the budding yeast: A comprehensive system to examine two-hybrid interactions in all possible combinations between the yeast proteins. Proc Natl Acad Sci USA. 2000, 97: 1143-1147. 10.1073/pnas.97.3.1143.PubMedPubMed CentralView ArticleGoogle Scholar
- Uetz P, Giot L, Cagney G, Mansfield TA, Judson RS, Knight JR, Lockshon D, Narayan V, Srinivasan M, Pochart P, et al: A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae. Nature. 2000, 403: 623-627. 10.1038/35001009.PubMedView ArticleGoogle Scholar
- Gavin AC, Bosche M, Krause R, Grandi P, Marzioch M, Bauer A, Schultz J, Rick JM, Michon AM, Cruciat CM, et al: Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature. 2002, 415: 141-147. 10.1038/415141a.PubMedView ArticleGoogle Scholar
- Ho Y, Gruhler A, Heilbut A, Bader GD, Moore L, Adams SL, Millar A, Taylor P, Bennett K, Boutilier K, et al: Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry. Nature. 2002, 415: 180-183. 10.1038/415180a.PubMedView ArticleGoogle Scholar
- Cho RJ, Campbell MJ, Winzeler EA, Steinmetz L, Conway A, Wodicka L, Wolfsberg TG, Gabrielian AE, Landsman D, Lockhart DJ, et al: A genome-wide transcriptional analysis of the mitotic cell cycle. Molecular Cell. 1998, 2: 65-73. 10.1016/S1097-2765(00)80114-8.PubMedView ArticleGoogle Scholar
- Kanehisa M, Goto S, Kawashima S, Okuno Y, Hattori M: The KEGG resource for deciphering the genome. Nucleic Acids Res. 2004, D277-280. 10.1093/nar/gkh063. 32 DatabaseGoogle Scholar
- Lee TI, Rinaldi NJ, Robert F, Odom DT, Bar-Joseph Z, Gerber GK, Hannett NM, Harbison CT, Thompson CM, Simon I, et al: Transcriptional regulatory networks in Saccharomyces cerevisiae. Science. 2002, 298: 799-804. 10.1126/science.1075090.PubMedView ArticleGoogle Scholar
- Horak CE, Luscombe NM, Qian J, Bertone P, Piccirrillo S, Gerstein M, Snyder M: Complex transcriptional circuitry at the G1/S transition in Saccharomyces cerevisiae. Genes Dev. 2002, 16: 3017-3033. 10.1101/gad.1039602.PubMedPubMed CentralView ArticleGoogle Scholar
- Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, Page N, Robinson M, Raghibizadeh S, Hogue CW, Bussey H, et al: Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science. 2001, 294: 2364-2368. 10.1126/science.1065810.PubMedView ArticleGoogle Scholar
- Nakaya A, Goto S, Kanehisa M: Extraction of correlated gene clusters by multiple graph comparison. Genome Inform Ser. 2001, 12: 44-53.Google Scholar
- Albert R, Barabasi AL: Statistical mechanics of complex networks. Rev Modern Phys. 2002, 74: 47-97. 10.1103/RevModPhys.74.47.View ArticleGoogle Scholar
- Amaral LA, Scala A, Barthelemy M, Stanley HE: Classes of small-world networks. Proc Natl Acad Sci USA. 2000, 97: 11149-11152. 10.1073/pnas.200327197.PubMedPubMed CentralView ArticleGoogle Scholar
- Girvan M, Newman ME: Community structure in social and biological networks. Proc Natl Acad Sci USA. 2002, 99: 7821-7826. 10.1073/pnas.122653799.PubMedPubMed CentralView ArticleGoogle Scholar
- Jeong H, Mason SP, Barabasi AL, Oltvai ZN: Lethality and centrality in protein networks. Nature. 2001, 411: 41-42. 10.1038/35075138.PubMedView ArticleGoogle Scholar
- Yu H, Zhu X, Greenbaum D, Karro J, Gerstein M: TopNet: a tool for comparing biological sub-networks, correlating protein properties with topological statistics. Nucleic Acids Res. 2004, 32: 328-337. 10.1093/nar/gkh164.PubMedPubMed CentralView ArticleGoogle Scholar
- Milo R, Shen-Orr S, Itzkovitz S, Kashtan N, Chklovskii D, Alon U: Network motifs: simple building blocks of complex networks. Science. 2002, 298: 824-827. 10.1126/science.298.5594.824.PubMedView ArticleGoogle Scholar
- Yeger-Lotem E, Sattath S, Kashtan N, Itzkovitz S, Milo R, Pinter RY, Alon U, Margalit H: Network motifs in integrated cellular networks of transcription-regulation and protein-protein interaction. Proc Natl Acad Sci USA. 2004, 101: 5934-5939. 10.1073/pnas.0306752101.PubMedPubMed CentralView ArticleGoogle Scholar
- Luscombe NM, Babu MM, Yu H, Snyder M, Teichmann SA, Gerstein M: Genomic analysis of regulatory network dynamics reveals large topological changes. Nature. 2004, 431: 308-312. 10.1038/nature02782.PubMedView ArticleGoogle Scholar
- Ihmels J, Levy R, Barkai N: Principles of transcriptional control in the metabolic network of Saccharomyces cerevisiae. Nat Biotechnol. 2004, 22: 86-92. 10.1038/nbt918.PubMedView ArticleGoogle Scholar
- Balazsi G, Barabasi AL, Oltvai ZN: Topological units of environmental signal processing in the transcriptional regulatory network of Escherichia coli. Proc Natl Acad Sci USA. 2005, 102: 7841-7846. 10.1073/pnas.0500365102.PubMedPubMed CentralView ArticleGoogle Scholar
- Guelzim N, Bottani S, Bourgine P, Kepes F: Topological and causal structure of the yeast transcriptional regulatory network. Nat Genet. 2002, 31: 60-63. 10.1038/ng873.PubMedView ArticleGoogle Scholar
- Fraser HB, Hirsh AE, Steinmetz LM, Scharfe C, Feldman MW: Evolutionary rate in the protein interaction network. Science. 2002, 296: 750-752. 10.1126/science.1068696.PubMedView ArticleGoogle Scholar
- Yu H, Greenbaum D, Xin Lu H, Zhu X, Gerstein M: Genomic analysis of essentiality within protein networks. Trends Genet. 2004, 20: 227-231. 10.1016/j.tig.2004.04.008.PubMedView ArticleGoogle Scholar
- Hinnebusch AG, Natarajan K: Gcn4p, a master regulator of gene expression, is controlled at multiple levels by diverse signals of starvation and stress. Eukaryot Cell. 2002, 1: 22-32. 10.1128/EC.01.1.22-32.2002.PubMedPubMed CentralView ArticleGoogle Scholar
- Drysdale CM, Duenas E, Jackson BM, Reusser U, Braus GH, Hinnebusch AG: The transcriptional activator GCN4 contains multiple activation domains that are critically dependent on hydrophobic amino acids. Mol Cell Biol. 1995, 15: 1220-1233.PubMedPubMed CentralView ArticleGoogle Scholar
- Ge H, Liu Z, Church GM, Vidal M: Correlation between transcriptome and interactome mapping data from Saccharomyces cerevisiae. Nat Genet. 2001, 29: 482-486. 10.1038/ng776.PubMedView ArticleGoogle Scholar
- Grigoriev A: A relationship between gene expression and protein interactions on the proteome scale: analysis of the bacteriophage T7 and the yeast Saccharomyces cerevisiae. Nucleic Acids Res. 2001, 29: 3513-3519. 10.1093/nar/29.17.3513.PubMedPubMed CentralView ArticleGoogle Scholar
- Kemmeren P, van Berkum NL, Vilo J, Bijma T, Donders R, Brazma A, Holstege FC: Protein interaction verification and functional annotation by integrated analysis of genome-scale data. Mol Cell. 2002, 9: 1133-1143. 10.1016/S1097-2765(02)00531-2.PubMedView ArticleGoogle Scholar
- Jansen R, Greenbaum D, Gerstein M: Relating whole-genome expression data with protein-protein interactions. Genome Res. 2002, 12: 37-46. 10.1101/gr.205602.PubMedPubMed CentralView ArticleGoogle Scholar
- Yu H, Luscombe NM, Qian J, Gerstein M: Genomic analysis of gene expression relationships in transcriptional regulatory networks. Trends Genet. 2003, 19: 422-427. 10.1016/S0168-9525(03)00175-6.PubMedView ArticleGoogle Scholar
- Shen-Orr SS, Milo R, Mangan S, Alon U: Network motifs in the transcriptional regulation network of Escherichia coli. Nature Genetics. 2002, 31: 64-68. 10.1038/ng881.PubMedView ArticleGoogle Scholar
- Zaslaver A, Mayo AE, Rosenberg R, Bashkin P, Sberro H, Tsalyuk M, Surette MG, Alon U: Just-in-time transcription program in metabolic pathways. Nat Genet. 2004, 36: 486-491. 10.1038/ng1348.PubMedView ArticleGoogle Scholar
- Kalir S, Alon U: Using a quantitative blueprint to reprogram the dynamics of the flagella gene network. Cell. 2004, 117: 713-720. 10.1016/j.cell.2004.05.010.PubMedView ArticleGoogle Scholar
- Basu S, Mehreja R, Thiberge S, Chen MT, Weiss R: Spatiotemporal control of gene expression with pulse-generating networks. Proc Natl Acad Sci USA. 2004, 101: 6355-6360. 10.1073/pnas.0307571101.PubMedPubMed CentralView ArticleGoogle Scholar
- Mangan S, Alon U: Structure and function of the feed-forward loop network motif. Proc Natl Acad Sci USA. 2003, 100: 11980-11985. 10.1073/pnas.2133841100.PubMedPubMed CentralView ArticleGoogle Scholar
- Spellman P, Sherlock G, Zhang M, Iyer V, Anders K, Eisen M, Brown P, Botstein D, Futcher B: Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol Biol Cell. 1998, 9: 3273-3297.PubMedPubMed CentralView ArticleGoogle Scholar
- Mazurie A, Bottani S, Vergassola M: An evolutionary and functional assessment of regulatory network motifs. Genome Biol. 2005, 6: R35-10.1186/gb-2005-6-4-r35.PubMedPubMed CentralView ArticleGoogle Scholar
- Wingender E, Chen X, Fricke E, Geffers R, Hehl R, Liebich I, Krull M, Matys V, Michael H, Ohnhauser R, et al: The TRANSFAC system on gene expression regulation. Nucleic Acids Res. 2001, 29: 281-283. 10.1093/nar/29.1.281.PubMedPubMed CentralView ArticleGoogle Scholar
- Hodges PE, McKee AH, Davis BP, Payne WE, Garrels JI: The Yeast Proteome Database (YPD): a model for the organization and presentation of genome-wide functional data. Nucleic Acids Res. 1999, 27: 69-73. 10.1093/nar/27.1.69.PubMedPubMed CentralView ArticleGoogle Scholar
- Bader GD, Betel D, Hogue CW: BIND: the Biomolecular Interaction Network Database. Nucleic Acids Res. 2003, 31: 248-250. 10.1093/nar/gkg056.PubMedPubMed CentralView ArticleGoogle Scholar
- Mewes HW, Frishman D, Guldener U, Mannhaupt G, Mayer K, Mokrejs M, Morgenstern B, Munsterkotter M, Rudd S, Weil B: MIPS: a database for genomes and protein sequences. Nucleic Acids Res. 2002, 30: 31-34. 10.1093/nar/30.1.31.PubMedPubMed CentralView ArticleGoogle Scholar
- Xenarios I, Salwinski L, Duan XJ, Higney P, Kim SM, Eisenberg D: DIP, the Database of Interacting Proteins: a research tool for studying cellular networks of protein interactions. Nucleic Acids Res. 2002, 30: 303-305. 10.1093/nar/30.1.303.PubMedPubMed CentralView ArticleGoogle Scholar
- von Mering C, Krause R, Snel B, Cornell M, Oliver SG, Fields S, Bork P: Comparative assessment of large-scale data sets of protein-protein interactions. Nature. 2002, 417: 399-403. 10.1038/nature750.PubMedView ArticleGoogle Scholar
- Jansen R, Yu H, Greenbaum D, Kluger Y, Krogan NJ, Chung S, Emili A, Snyder M, Greenblatt JF, Gerstein M: A Bayesian networks approach for predicting protein-protein interactions from genomic data. Science. 2003, 302: 449-453. 10.1126/science.1087361.PubMedView ArticleGoogle Scholar
- Supplementary Data Website. [http://networks.gersteinlab.org/network/netcomp/]
- Zhang LV, King OD, Wong SL, Goldberg DS, Tong AH, Lesage G, Andrews B, Bussey H, Boone C, Roth FP: Motifs, themes and thematic maps of an integrated Saccharomyces cerevisiae interaction network. J Biol. 2005, 4: 6-10.1186/jbiol23.PubMedPubMed CentralView ArticleGoogle Scholar
- Qian J, Dolled-Filhart M, Lin J, Yu H, Gerstein M: Beyond synexpression relationships: local clustering of time-shifted and inverted gene expression profiles identifies new, biologically relevant interactions. J Mol Biol. 2001, 314: 1053-1066. 10.1006/jmbi.2000.5219.PubMedView ArticleGoogle Scholar
- Cupp JR, McAlister-Henn L: Kinetic analysis of NAD(+)-isocitrate dehydrogenase with altered isocitrate binding sites: contribution of IDH1 and IDH2 subunits to regulation and catalysis. Biochemistry. 1993, 32: 9323-9328. 10.1021/bi00087a010.PubMedView ArticleGoogle Scholar
- Iyer VR, Horak CE, Scafe CS, Botstein D, Snyder M, Brown PO: Genomic binding sites of the yeast cell-cycle transcription factors SBF and MBF. Nature. 2001, 409: 533-538. 10.1038/35054095.PubMedView ArticleGoogle Scholar
- Daignan-Fornier B, Fink GR: Coregulation of purine and histidine biosynthesis by the transcriptional activators BAS1 and BAS2. Proc Natl Acad Sci USA. 1992, 89: 6746-6750. 10.1073/pnas.89.15.6746.PubMedPubMed CentralView ArticleGoogle Scholar
- Dirick L, Bohm T, Nasmyth K: Roles and regulation of Cln-Cdc28 kinases at the start of the cell cycle of Saccharomyces cerevisiae. EMBO J. 1995, 14: 4803-4813.PubMedPubMed CentralGoogle Scholar
- Machado AK, Morgan BA, Merrill GF: Thioredoxin reductase-dependent inhibition of MCB cell cycle box activity in Saccharomyces cerevisiae. J Biol Chem. 1997, 272: 17045-17054. 10.1074/jbc.272.27.17045.PubMedView ArticleGoogle Scholar
- Matangkasombut O, Buratowski S: Different sensitivities of bromodomain factors 1 and 2 to histone H4 acetylation. Mol Cell. 2003, 11: 353-363. 10.1016/S1097-2765(03)00033-9.PubMedView ArticleGoogle Scholar
- Miyoshi K, Shirai C, Mizuta K: Transcription of genes encoding trans-acting factors required for rRNA maturation/ribosomal subunit assembly is coordinately regulated with ribosomal protein genes and involves Rap1 in Saccharomyces cerevisiae. Nucleic Acids Res. 2003, 31: 1969-1973. 10.1093/nar/gkg278.PubMedPubMed CentralView ArticleGoogle Scholar
- Ogawa N, Noguchi K, Sawai H, Yamashita Y, Yompakdee C, Oshima Y: Functional domains of Pho81p, an inhibitor of Pho85p protein kinase, in the transduction pathway of Pi signals in Saccharomyces cerevisiae. Mol Cell Biol. 1995, 15: 997-1004.PubMedPubMed CentralView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.