(516n) A Systematic Framework For Modeling Transcriptional Regulatory Networks

Reverse engineering transcriptional regulatory networks is a major task in computational biology. Knowledge about transcriptional regulation is a step towards gaining insights into diseases since many diseases are characterized by abnormal gene regulation. Transcription initiation is affected by the presence of specific DNA-binding proteins (known as transcription factors ? TFs) whose fundamental role is to bind to the promoter region of candidate target genes and thereby control their expression. Hence, cells adjust gene expression profiles in response to environmental changes through the activation of a series of signal transduction pathways which in turn modulate the activities of several transcription factors [1].

Regulator transcription levels are generally not appropriate measures of transcription factor activity [2]. Recently, methods combining TF-gene connectivity data and gene expression measurements have emerged in order to quantify these regulatory interactions [3-11]. The main goal of this reverse engineering is to identify the activation program of transcription modules under particular conditions [12] so as to hypothesize how activation/deactivation of gene expression can be induced/suppressed [13]. Aside from the development of descriptive models that correlate TFA and expression of target genes, a critical question becomes how to identify those TFs that significantly contribute to regulation and should be modulated. Along those lines the authors [14] speculate that the mRNA profile of the target gene should be similar to the reconstructed TFA for the regulating proteins, in [15] they claim that accurate binding information should lead to robust TFA reconstructions whereas in [16] develop a greedy-based selection of critical regulators.

In this study we propose an optimization ? based model (MILP formulation) that aims at identifying the optimal reconstruction and architectures in a rigorous manner. We propose a systematic construction of alternative regulatory structures in parallel with a consistency metric for assessing the robustness of specific transcription factors. The goal is to develop a model that can provide meaningful biological insights on gene regulation. Our primary aim is to identify the best possible regulatory decomposition while utilizing prior biological knowledge about known interactions in terms of existence as well as in terms of known directionality (activation/suppression). Furthermore, another key aspect of our model is that we can also infer the regulatory role of those transcription factors that their activity on certain promoter regions is unknown ? it can be either activation or repression. Our model (miSARN) allows us to identify in a systematic and rigorous way alternative regulatory architectures, which will unravel the existence of a set of robust and presumably critical regulators. Such identification might serve as a diagnostic tool for in silico target identification [17].

We validated our model on temporal expression profiles of E coli during transition from glucose to acetate as the sole carbon source. This dataset consists of the measured expression values of 100 genes recorded at 10 time points. Such expression data have been part of previous studies [18-21] . The corresponding connectivity matrix is extracted from the available information of RegulonDB [22] database. We identified a set of critical regulators based on their robust profiles while emphasizing their biological relevance to specific experimental conditions. Also, our model can generate a number of equivalent regulatory structures that play a major role in explaining the viability of different strains of E coli as well as its ability to tolerate a variety of environmental conditions while still retaining functionality.

1. Li H, Wang W: Dissecting the transcription networks of a cell using computational genomics. Curr Opin Genet Dev 2003, 13:611-616. 2. Niederbichler AD, Hoesel LM, Westfall MV, Gao H, Ipaktchi KR, Sun L, Zetoune FS, Su GL, Arbabi S, Sarma JV, et al: An essential role for complement C5a in the pathogenesis of septic cardiac dysfunction. J Exp Med 2006, 203:53-61. 3. Alter O, Golub GH: Integrative analysis of genome-scale data by using pseudoinverse projection predicts novel correlation between DNA replication and RNA transcription. Proc Natl Acad Sci U S A 2004, 101:16577-16582. 4. Kato M, Hata N, Banerjee N, Futcher B, Zhang MQ: Identifying combinatorial regulation of transcription factors and binding motifs. Genome Biol 2004, 5:R56. 5. Gao F, Foat BC, Bussemaker HJ: Defining transcriptional networks through integrative modeling of mRNA expression and transcription factor binding data. BMC Bioinformatics 2004, 5:31. 6. Boulesteix AL, Strimmer K: Predicting transcription factor activities from combined analysis of microarray and ChIP data: a partial least squares approach. Theor Biol Med Model 2005, 2:23. 7. Yeung MK, Tegner J, Collins JJ: Reverse engineering gene networks using singular value decomposition and robust regression. Proc Natl Acad Sci U S A 2002, 99:6163-6168. 8. Bussemaker HJ, Li H, Siggia ED: Regulatory element detection using correlation with expression. Nat Genet 2001, 27:167-171. 9. Kao KC, Tran LM, Liao JC: A global regulatory role of gluconeogenic genes in Escherichia coli revealed by transcriptome network analysis. J Biol Chem 2005, 280:36079-36087. 10. Tran LM, Brynildsen MP, Kao KC, Suen JK, Liao JC: gNCA: A framework for determining transcription factor activity based on transcriptome: identifiability and numerical implementation. Metabolic Engineering 2005, 7:128-141. 11. Sun N, Carroll RJ, Zhao H: Bayesian error analysis model for reconstructing transcriptional regulatory networks. Proc Natl Acad Sci U S A 2006, 103:7988-7993. 12. Wang W, Cherry JM, Botstein D, Li H: A systematic approach to reconstructing transcription networks in Saccharomycescerevisiae. Proc Natl Acad Sci U S A 2002, 99:16893-16898. 13. Ng A, Bursteinas B, Gao Q, Mollison E, Zvelebil M: pSTIING: a 'systems' approach towards integrating signalling pathways, interaction and transcriptional regulatory networks in inflammation and cancer. Nucleic Acids Res 2006, 34:D527-534. 14. Gao F, Foat BC, Bussemaker HJ: Defining transcriptional networks through integrative modeling of mRNA expression and transcription factor binding data. BMC Bioinformatics 2004, 5:31. 15. Sun N, Carroll RJ, Zhao H: Bayesian error analysis model for reconstructing transcriptional regulatory networks. Proc Natl Acad Sci U S A 2006, 103:7988-7993. 16. Chen KC, Wang TY, Tseng HH, Huang CY, Kao CY: A stochastic differential equation model for quantifying transcriptional regulatory network in Saccharomyces cerevisiae. Bioinformatics 2005, 21:2883-2890. 17. Sun N, Carroll RJ, Zhao H: Bayesian error analysis model for reconstructing transcriptional regulatory networks. Proc Natl Acad Sci U S A 2006, 103:7988-7993. 18. Boulesteix AL, Strimmer K: Predicting transcription factor activities from combined analysis of microarray and ChIP data: a partial least squares approach. Theor Biol Med Model 2005, 2:23. 19. Kao KC, Yang YL, Boscolo R, Sabatti C, Roychowdhury V, Liao JC: Transcriptome-based determination of multiple transcription regulator activities in Escherichia coli by using network component analysis. Proc Natl Acad Sci U S A 2004, 101:641-646. 20. Pournara I, Wernisch L: Factor analysis for gene regulatory networks and transcription factor activity profiles. BMC Bioinformatics 2007, 8:61. 21. Drazinic CM, Smerage JB, Lopez MC, Baker HV: Activation mechanism of the multifunctional transcription factor repressor-activator protein 1 (Rap1p). Mol Cell Biol 1996, 16:3187-3196. 22. Salgado H, Santos-Zavaleta A, Gama-Castro S, Millan-Zarate D, Diaz-Peredo E, Sanchez-Solano F, Perez-Rueda E, Bonavides-Martinez C, Collado-Vides J: RegulonDB (version 3.2): transcriptional regulation and operon organization in Escherichia coli K-12. Nucleic Acids Res 2001, 29:72-74.