GENERAL STATEMENT |
My research focuses on applying systems theory to challenging problems related to modeling of complex chemical reaction networks, including two parts: 1) Mathematical modeling for signal transduction pathways in the field of Systems Biology; 2) Sensor location in a Binary Distillation Column for chemical plants. Before I joined Chemical Engineering Department, I did some research on process control for thermal systems in my Master thesis.
|
MOTIVATION |
- The investigation of signal transduction pathways is one of the central themes in systems biology
- Signal transduction regulates many cellular processes and is also involved in extracellular communication
- Malfunctioning of signal transduction pathways can lead to diseases like certain types of cancer (e.g. JAK/STAT is associated with colon cancer and failure of MAPK with gastric cancer)
- Mathematical modeling and analysis plays an important role for studying signal transduction mechanism in systems biology, as it can
- improve understanding of signal transduction mechanisms
- predict cellular response
- indicate potential treatment options
- The mathematical model has to be updated and verified by quantitative data before it can be put to wide-spread use. However, modeling of signal transduction pathways is characterized by a lack of quantitative data. To address this challenge, my Ph.D. research focuses on how to derive quantitative data for transcription factors (the important components in the reaction network) and how to model for the situation where only limited quantitative data is available.
|
OUTLINE OF THE RESEARCH |
- My interdisciplinary research includes:
- Project 1: Deriving quantitative data for transcription factors by analyzing GFP images and solving an inverse problem.
- Project 2: Integrating available qualitative information into modeling by using Fuzzy logic.
- Project 3: Performing nonlinear model reduction of uncertain systems.
- Project 4: Sensor network and experimental design for chemical plants
- Project 1 presents an integrated modeling and experimental approach to derive the quantitative data for transcription factors. Based on the quantitative data obtained from Project 1, Project 2 and 3 address two different solutions to the challenging problem how to model when limited quantitative data is available. Different from the subject in Project 1 ~ 3, i.e., modeling for signaling pathways in the field of systems biology, Project 4 deals with how to design sensor networks in a Binary Distillation Column for chemical plants.
|
PROJECT 1: Deriving quantitative data for transcription factors by analyzing GFP images and solving an inverse problem |
- Modeling in systems biology is characterized by limited quantitative data. Based on a GFP reporter system, an integrated modeling and experimental approach for measuring transcription factor profile has been developed:
- Step1: GFP fluorescence images are obtained from the GFP reporter system stimulated by cytokine (e.g. TNF-α, IL-6, and IL-10);
- Step2: Image analysis technique is performed to extract fluorescent intensity profiles from GFP images;
- Image analysis based on Principal Component Analysis (PCA) and K-means clustering
- Image analysis based on Wavelet De-noising and mathematical morphology
Step3: An inverse problem is solved that infers the transcription profile from the fluorescent intensity profile making use of a model describing GFP transcription;
Step4: The quantitative data is used for data-driven modeling to validate or refine existing models.
- Quantitative data for STAT3N*-STAT3N*, C/EBPβ, and NF-κB have been obtained.
- Please refer to the BMC paper and the IET paper for details.
|
PROJECT 2: Integrating available qualitative information into modeling by using Fuzzy logic |
- Many systems investigated in systems biology are characterized by a lack of quantitative data, yet a significant amount of qualitative knowledge is available. This is a situation which seems ideally suited for fuzzy modeling as the qualitative information can be incorporated into the model building process in the form of linguistic rules, while only parameters of membership functions need to be estimated from (the limited amount of) available data. A Fuzzy modeling procedure has been proposed for signaling pathways. Please refer to the Chemical Engineering Science paper for details
|
PROJECT 3: Performing nonlinear model reduction of uncertain systems |
- Validating fundamental models for signaling pathway is nontrivial due to: 1) fundamental models contain a large number of uncertain parameters; 2) not all the components in the model can be assessed by experiment; 3) the quantitative data is limited. One option to address this point is to perform model reduction. A model reduction technique that keeps the physical meaning of some of the states and parameters of the model has been investigated which makes use of sensitivity analysis and parameter clustering. The model reduction technique is applied to reduce a IL-6 signaling pathway from an ODE model with 66 states and 115 parameters to a simplified model with only 11 states and 16 parameters.
|
PROJECT 4: Sensor network and experimental design for chemical plants |
- This work deals with the determination of where sensors should be placed in a chemical plant to obtain as much information about the plant behavior as possible. The technique also finds application in experimental design as one wants to determine which quantities should be measured so that as much information about the non-measured variables can be obtained. The proposed technique has been applied to a Binary Distillation Column.
|
PROJECTS participated before joining Chemical Engineering |
- Continuous wave laser photo-acoustic microscopy, Department of Electrical Engineering at Texas A&M University in College Station, 08/2005 ~ 05/2006. I was knowleged in this paper.
- Simultaneous identification of system order and parameters based on rational fraction equality, supervised by Professor Feng Ding in Department of Automation, Tsinghua University, 09/2001 ~ 11/2002.
- Identification of Output Error Model Based on Impulse Response, supervised by Professor Feng Ding in Department of Automation, Tsinghua University, 09/2001 ~ 11/2002.
- Gain scheduled control method and its application on the objects in thermal system, supervised by Professor Xuezhi Jiang and Professor Donghai Li in the Institute of Simulation & Control for Thermal Power Engineering, Tsinghua University, 11/2002 ~ 07/2004.
- The study of DCS control system for circulating fluidized-bed boiler, supervised by Professor Xuezhi Jiang and Professor Donghai Li in the Institute of Simulation & Control for Thermal Power Engineering, Tsinghua University, 11/2002 ~ 07/2004.
- A computer-network-based system to control and simulate the heating system in a 3-storeyed building with heat charging subsystem, supervised by Professor Xuezhi Jiang in the Institute of Simulation & Control for Thermal Power Engineering, Tsinghua University, 11/2002 ~ 07/2004.
- Experimental Study on the Character of Convective Heat Transfer of the High-moisture Flue Gas, supervised by Professor Yanguo Zhang in the Department of Thermal Engineering, Tsinghua University, 09/2000 ~ 07/2001.
|
