Introduction
- Systems Biology
- Bio-Inspired Engineering
- Research Interests
Research Focuses
- Signal Transduction Pathways
- Biomolecular Regulatory Networks
- Complex Cellular Dynamics
- Brain Systems Biology
- Cancer Systems Biology
- Cardiac Systems Biology
- Network Systems Biology
- Pregnosis
- Bio-inspired Self-Repairing Electronic Circuits
- Brain-inspired Speech Recognition System
Funded Research Projects
Software
- SBIE Software
- Source Codes of Publications
Research Equipments
Useful Information for Systems Biological Research and Links
Research Focuses
Systems Biology


Cellular Signal Transduction Pathways
Cells use receptors to continually monitor their environment and alter behavior accordingly. In order to determine how cells behave and interact, we need to understand how information is transferred both amongst and within cells. Cell signaling or "signal transduction" is the mechanism by which this transfer of biological information takes place. We have concentrated on signal- and systems-oriented approaches to investigate this flow of information, focusing specifically on the NF-kappaB, ERK, Wnt, JAK-STAT, and beta-AR pathways. Our aim was to identify the functional role of those pathways through quantitative relationships between parameters and variables. Since most of these relationships are nonlinear with varying feedback connections, this problem is non-trivial. In the future we plan to further investigate various crosstalks between pathways and apply the results to the problem of drug target identification.


Biomolecular Regulatory Networks
In order to probe complex biomolecular regulatory networks, we believe that a useful approach is to study their interactive structures. Therefore, inference of gene regulatory networks (GRN) for a specific subsystem or an entire genome scale can help us to unravel gene interaction mechanisms. Furthermore, we believe that we can utilize this information to identify novel drug targets and predict potential adverse effects. Thanks to the recent development of high-throughput measurement technologies, there is a renewed interest in unraveling hidden GRNs and various reverse engineering methods have been developed to infer them. However, due to both experimental limitations and methodological complexities, the majority of these attempts have not been completely successful. We have been interested in developing new methods by exploiting dynamical properties of temporal expression profiles and will further substantiate those properties through practical case studies.


Complex Cellular Dynamics
Cellular networks are composed of complex interconnections in which some subnetworks with particular functions can be identified as network motifs. Feedback loops, in particular, have been identified as playing critically important roles in regulating cellular behavior. Intriguingly, such feedback loops are often found as coupled structures in many cellular circuits. Using the available biological information, we have investigated the properties of coupled feedbacks and identified a set of three principles. Firstly, positive feedbacks enhance signal amplification and lead to bistable characteristics.  Secondly, negative feedbacks enhance homeostasis. Lastly, positive and negative feedbacks enable reliable decision making by properly modulating signal responses and effectively dealing with noise. Examples include apoptosis decision circuits, circadian regulatory circuits, and cellular memory circuits. We will further investigate various cellular dynamics to uncover their hidden design principles and extend them for bio-medical applications.


Brain Systems Biology
Brain functions (e.g., consciousness, creative thinking, memory formation, and recognition) are supposed to be mainly determined by the topological structure of neural networks composed of hundreds of thousands of neurons and their synaptic interactions. Recent studies suggested that network properties emerging from neural networks (e.g., synchronized oscillation between distant neural assemblies) can help us to uncover the underlying mechanisms of brain function although their correlations are largely unknown. Moreover, on the assumption that the selective synchronization is the basis for normal brain functioning, synchrony disruption could cause functional abnormalities such as epilepsy, Parkinson’s disease, schizophrenia, and so on. We are interested in such correlations between brain functions or disorders and emergent properties of relevant neural networks. In order to answer the question on what special network structures make them interlinked, we develop computational models of neural networks and identify the characteristics of network motifs having significant influence on the dynamics of the whole network. The ultimate goal is to investigate the emergent properties of a neural network in connection with brain functions and disorders, and unravel the hidden underlying mechanisms.


Cancer Systems Biology
Colorectal cancer originates from the rapidly renewing epithelium that covers the luminal surfaces of the large intestine and lines the colonic crypts. At the bottom of these crypts, stem cells continuously proliferate and produce transit-amplifying cells that rapidly divided several times before differentiating into the various cell types (e.g., absorptive columnar cells, Goblet cells, and neuroendocrine cells). Colorectal cancer arises when genetic alterations deregulate normal crypt dynamics. Proliferating cells are then no longer confined to the crypt’s base and the associated proliferative excess generates biomechanical stress on the crypt structure, which may deform in order to accommodate the newborn cells. In this study, we focus on understanding how deregulation of cell proliferation leads to the loss of homeostasis and malignant transformation of crypt dynamics. We also have interests in the identification of potential therapeutic interventions. To this end, we are developing a multi-scale platform model for colon crypt dynamics that integrates subcellular signaling networks (e.g., Wnt, BMP, Eph/Ephrine, Hedgehog signaling pathways, etc), the cell cycle clock and cell population dynamics of stem cells, transit-amplifying cells and differentiated cells.


Cardiac Systems Biology
Heart disease is one of the leading causes of death throughout the world. However, as cardiac functions are dynamically and complicatedly regulated, the underlying disease mechanisms of the heart have not yet been fully understood. In order to investigate complex cardiac disease mechanisms, we are currently developing a platform model that includes a large-scale signaling network related to cardiac hypertrophy. This platform model will help us to explore the development of pathological hypertrophy that often results in lethal heart diseases such as heart failure and arrhythmias using a systems-approach based on dynamical analysis of mathematical models. Moreover, we will establish drug/gene therapeutic strategies (multi-target drug & combinatorial therapies) in order to suppress or delay the development of pathological hypertrophy based on our platform model. Ultimately, we hope that the results obtained from our research contribute to the development of systematic methodologies to predict, diagnose, and treat various heart diseases.


Network Systems Biology
A large amount of information on bio-molecular interactions has been accumulated from advanced experimental technologies and omics-data analysis technologies. Through integrating such interaction information from various databases and literatures, we construct large-scale biological networks. The function of such biological networks is too complex to understand using traditional mathematical modeling techniques. Hence, instead of analyzing network dynamics by constructing a whole mathematical model, we explore dynamics and functions of networks by analyzing the topological properties such as network motifs, node degree distribution, clustering coefficient, path length, connection density, centrality, robustness, and modularity. The aim of network systems biology is to investigate the relations between topological properties and biological function.




Bio-inspired Engineering


Pregnosis
Pregnosis is a combination of the words “predictive” and “diagnosis” (created by Prof. Kwang-Hyun Cho). If we can diagnose critical events before they actually occur, we can signal “early warning” and prevent them. Complex dynamical systems, such as biological systems, ecosystems, climate, and man-made complex systems such as manufacturing systems and financial markets, can have tipping points at which a sudden shift to a contrasting dynamical regime may occur (M. Scheffer et al., “Early-warning signals for critical transitions,” Nature Reviews, vol. 461, pp. 53-59, Sep. 2009). Although predicting such critical points before they are reached is difficult, studies in different scientific areas suggested the existence of generic early-warning signals. A critical transition means that a system shifts abruptly from one state to another and it seems true that there are such generic early-warning signals before the occurrence of critical transitions regardless of the types or internal dynamics of complex dynamical systems. These generic early-warning signals can be explained based on attractor dynamics. An attractor indicates a stable state to which a system state ultimately converges. For each attractor in a multiple-attractor system, the perturbed states within a certain range around an attractor eventually converge to the attractor. Such a range around the attractor is called the basin of attraction. We are investigating the precursor phenomena seen as getting close to a bifurcation point which triggers a critical transition from one to another attractor state to find out novel early-warning signals for practical applications.


Bio-inspired Self-Repairing Electronic Circuits
Through millions of years of evolution, biological systems have developed extremely high level of reliability even though they seem to be organized just complicatedly. Our body composed of trillions of cells can recover even a strong damage by itself. If we look at them closely, we can find that there are many hidden recovery processes which help cells repair the damages by themselves. For instance, at a molecular level in the cell, several DNA repair processes constantly recover DNA damages, though DNA is having about 1 million damages per cell everyday. Even if the number of DNA damages exceeds the capabilities of DNA repair processes, the cell with those damaged DNA dies through apoptosis, then it triggers a stem cell to differentiate into the same functioning cell as the dead cell by expressing the corresponding genome. The goal of our study in this direction is to investigate the underlying mechanism of self-repairing processes in living cells from the view of systems biology and then apply it to the development of self-repairing electronic circuits. The developed hardware structures with the self-repairing features can be widely used for various industrial applications as well as for some critical applications such as spaceship, flight, and nuclear plant where even a minor electronic device failure can cause serious risk.


Brain-inspired Speech Recognition System
Over the past several decades, automatic speech recognition (ASR) has made a huge progress; speech recognition accuracy in a quiet environment is now up to 95% these days. However, the recognition accuracy drops to 70% when noise is introduced. Even the state-of-the-art ASR system struggles with noisy environments. Many engineers have made a lot of efforts to overcome such a limitation so far, but they have not found any breakthrough yet. To make a breakthrough, we need an innovative idea. To challenge this problem, we are focusing on how our brain processes noisy speech signals. Recent technical improvements made it possible to measure brain activities with precision during speech comprehension, and various experimental measurements as well as new discoveries on how our brain deals with noisy speech signals are being updated. We follow the latest findings and investigate the newly discovered speech processing mechanism through mathematical modeling and simulations. Then we develop a novel bio-inspired ASR system and further apply it to an extensive set of speech samples. Our bio-inspired ASR system shows a higher recognition accuracy that is much robust to noise, which also partially explains the design principle of our brain for robust speech recognition.