Text Box: We are developing a parallel global optimization algorithm that can explore the protein energy landscape effectively and find the global energy minimum that corresponds to the protein native state. The algorithm combines local optimization with normal mode guided global search. Text Box: Text Box: Text Box: Parallel energy minimization Text Box: We are interested in developing a multi-scale method for protein modeling so that simulation of protein motion or optimization of protein conformation can be conducted more efficiently and accurately at a residual level with the force field provided through atomic level simulations. Text Box: Text Box:

Mutliscale protein modeling
Text Box: Text Box: Text Box:
Text Box: We are interested in developing a group of geometric databases for the retrieval of various geometric properties (and in particular, their distributions) in known proteins, such as distances, angles, volumes, etc., that may be useful for structural analysis, classification, as well as prediction.

Protein geometry databases
Text Box: Metabolic network modeling and pathway identification are important research subjects in systems biology. We are interested in developing efficient algorithms for large-scale steady-state metabolic network optimization, flux balancing, and pathway identification.

Metabolic network optimization
Text Box: More research related in Research Group Text Box: Iowa State University Text Box: DEPARTMENT OF MATHEMATICS

My general research area is in numerical optimization, which extends naturally to some areas of numerical linear algebra and numerical solutions to ordinary and partial differential equations. It is related to discrete optimization as well for many optimization problems have integer variables and are combinatorial in nature. My recent work, however, has been focused on applications of optimization (and related numerical and combinatorial subjects) in biology and in particular, in bio-molecular and bio-systems modeling. The following are the projects I am working on with my colleagues and students.

Research

We are developing efficient geometric buildup algorithms for the determination of protein structures with given sets of distance constraints, and also working on using optimization techniques and interval methods to obtain approximate structural ensembles for distance ranges.

Distance Based Protein Modeling

We are developing methods for refining low resolution protein structures obtained either experimentally or theoretically, by using additional conformational constraints such as inter-atomic distance constraints that can be extracted from databases of known protein structures.

Protein Structure Refinement

We are developing efficient and accurate numerical algorithms for simulating the transition of protein conformation from one state to another, with potential applications in the study of prion transformation that causes the Mad Cow Diseases. We are using a so-called multiple shooting approach for the problem.