GENERAL STATEMENT |
My goal in my research is: 1) My research can provide some potential solutions for the challenging problems we are facing in the modeling and analzying of complicated chemical reaction networks; 2) Other researchers can benefit from my work. Therefore, I am glad to share my work with people. My multidisciplinary work can be implemented in systems biology, process systems engineering, pattern recognition, mathematical modeling for chemical reaction network, and uncertainty modeling. I am now supervising two undergraduate students to clean up my codes. I will put them onto my advisor's webpage some time later. Please send me email if you want to get my code.
|
IMAGE ANALYSIS PROGRAMS |
- Image analysis based on PCA and K-means clustering
- This image analysis method distinguishes regions of the image with similar features, e.g., brightness，and then clusters the pixels into different groups which correspond to different levels of the feature.
- This method can be applied to recognize the objects from the noisy background. This approach can be applied to monitor the partical size distribution for crystallization process in specialty chemical, pharmaceutical industry.
|
MATHEMATICAL MODELING PROGRAMS FOR CHEMICAL REACTION NETWORKS |
- Derive the ODE model from chemical reaction networks
- Reactions with known kinetic parameters, e.g. A + B -> C + D with the rate constant k
- Reactions in form of differential equations, e.g., d[A]/dt = -k [A]
- Derive the ODE model from the reactions in the above two forms
- Derive the uncertainty of the estimated parameters from the uncertainty of the experimental data
- Assume the uncertainty of the experimental data follows Gaussian distribution
- Apply Markov chain Monte Carlo (MCMC) to sample new data sets from the experimental data
- Estimate parameters for each new data set
- Calculate the uncertainty of the estimated parameters
|
SENSITIVITY ANALYSIS FOR THE PARAMETERS IN THE MODELS OF CHEMICAL REACTION NETWORKS |
- The program shows which parameters have the most important impact on the system behavior.
- The information from sensitivity analysis plays a key role on parameter estimation as it determines which parameters should be estimated when only limited experimental data is available.
|
MATHEMATICAL MODELS FOR SIGNAL TRANSDUCTION PATHWAYS |
- The mathematical models developed in my Ph.D. research are mainly involved in accute phase response of inflammation or non-alcoholic fatty liver disease (NAFLD). They can be applied in drug discovery, especially in the target identification. The modeling techniques, such as selecting the model structure, building the ODE models, analyzing the model to locate the important components in the signaling pathway, and parameter estimation, can be applied in modeling for other signal transduction pathways in the field of Systems Biology.
- Mathematical model for TNF-a ~ NFκB signal transduction pathway
- The ODE model consists of 37 states and 62 parameters. Program can be downloaded from Dr.Hahn's webpage.
- Mathematical model for IL-6 ~ Jak/STAT + Erk-C/EBPβ signal transduction pathway
- The ODE model consists of 66 differential equations and 115 parameters. Program can be downloaded from Dr.Hahn's webpage
- Mathematical model for IL-6 + IL-10~ Jak/STAT + Erk-C/EBPβ signal transduction pathway (under construction)
- The ODE model consists of 77 differential equations and 128 parameters. This model can be applied to investigate the potential treatment options for steatosis.
|
SOME USEFUL LINKS |
- Professor Juergen Hahn Group
- Process Systems Engineering Community in USA: CEPAC
- Journal in the field of Process System Engineering, with a specific emphasis on Chemical Engineering
- Journal in the field of Systems Biology
|
