Research Summary

The research in my lab is aimed at a systems level quantitative understanding of the structure, function, and evolution of gene regulatory networks. Our goals are driven by the following questions: 1) How does cell process information about the environment and optimize its global response? What is it optimizing? What are the basic constraints? What are the design principles that will emerge from the optimization? 2) How do different regulatory systems interact with each other? What emerging properties will we uncover as we attempt an integration? 3) How does evolution constrain network structure or vise versa? What are the possible evolutionary paths that connect different structures? What is the functional importance of large scale rewiring?

To address these questions, we take an interdisciplinary approach that combines exquisitely quantitative experimentation with theoretical modeling and bioinformatic analysis. Our research program consists mainly of the following three components:
1) to accurately reconstruct the cellular transcription networks using functional genomics and bioinformatics tools;
2) to study the dynamics of gene regulatory networks quantitatively at a systems level;
3) to derive principles governing the evolution of gene regulatory networks.
These components are intimately connected and they inform each other. The first component is aimed at elucidating the structure of the network, which forms the basis for the evolutionary and functional analysis. The second component seeks to understand the functional constraints imposed by the dynamic performance of the cell in a fluctuating environment, and the third component addresses fundamental evolutionary constraints.

1) network reconstruction
We have developed a number of computational algorithms to reconstruct the transcription networks of a cell by combining sequence information with gene expression data based on bioinformatics analysis, and have employed mechanistic models to study the dynamic behavior of the reconstructed networks.  We have developed computational algorithms for identifying regulatory elements in a genome based on statistical analysis of sequence and gene expression data  (Bussemaker et al. PNAS 2000; Bussemaker et al. Nat. Gen. 2001; Li et al. PNAS 2002, Wu et al. BMC Bioinformatics 2007, Grskovic et al. PLoS genetics 2007), algorithms for identifying the target genes of transcription factors, and algorithms for inferring the conditions under which each factor is activated or repressed (Wang et al. PNAS 2002; Wang et al. PNAS 2005). We have developed systematic approaches for analyzing combinatorial regulation by co-activated transcription factors, in order to understand how combinations of binding sites in a promoter can define complex regulatory functions (Wang et al. PNAS 2005).  We have applied these tools to analyzing specific biological systems in collaboration with experimental labs (Murphy et al. Nature 2003; Patil et al. PLoS biology, 2004; McCarroll et al. Nature Gen. 2004; Jolly et al. BMC bioinformatics 2005).

2) dynamics
To understand the functional constraints and design principles of gene regulatory networks, we quantitatively study the dynamics of the network’s response to environment/genetic perturbations. We have recently developed an automated system to study protein expression with high accuracy and high temporal resolution, for a number of genes simultaneously, at the single cell level. We have used this system to study the dynamics and design principles of a basic regulatory architecture commonly found in many metabolic pathways and discovered that combinatorial regulation of enzymes on a linear pathway leads to differential responses that optimize the speed of nutrient recovery instead of the recovery of the steady state level. We are currently developing a high throughput system to quantitative measure proteome-wide expression in single cells with high temporal resolution.

3) evolution
To derive principles governing the evolution of gene regulatory networks, we are beginning to systematically characterize different transcriptional circuits in fungi species in collaboration with Sandy Johnson’s lab. We have analyzed the evolution of the transcriptional circuit controlling yeast mating types. We discovered a mechanism by which a transcriptional circuit can be completely rewired while maintaining the some logic throughout evolution (Tsong et al, Nature 2006). To generalize what we have learned from the mating type control, we have recently analyzed a larger transcriptional circuitry involving the transcription factor MCM1 and several of its co-factors, responsible for regulating a range of systems from mating type to cell cycle. We again found large scale rewiring with functional conservation. Most interestingly, we discovered that a large group of genes with similar function can be recruited to the MCM1 circuit, probably through the evolution of new combinatorial interaction between MCM1 and other co-factors, suggesting that these changes are adaptive, and that combinatorial interaction may facilitate the rewiring of a large transcriptional circuit (Tuch et al, PLoS Biology 2008).