Welcome to the Su Lab

• Computational inference and modeling of regulatory networks/pathways

The availability of increasing number of sequenced genomes and other forms of high throughput data sets has provided unprecedented opportunities for computational inference of regulatory networks/pathways underlying basically all biological processes in a cell. We are committed to develop a computational framework and a suite of software for inference of regulatory networks/pathways through mining various types of data, and apply these tools to construction of important regulatory networks/pathways in both prokaryotic and eukaryotic cells.

•  Characterization and modeling of transcriptional regulatory networks in prokaryotic and eukaryotic cells

Identification of transcriptional regulatory networks in both prokaryotic and eukaryotic cells remains a grand challenging task for both computational and experimental biology in this post-genome era.  We are particularly interested in developing an integrated approach of computational prediction and experimental investigation (1) to identify all cis-regulatory elements in a genome, (2) to map each identified cis-regulatory element to its cognate DNA bind protein, and (3) to construct transcriptional regulatory networks in a prokaryotic cell or in a specific eukaryotic cell type using the predicted cis-regulatory elements, the mapping between cis-regulatory elements and DNA binding proteins, and time-series microarrary gene expression data.

•  Molecular modeling and simulation of protein/membrane complexes

 Membrane proteins comprise up to 30% of the proteome of a genome, and they are involved in a broad region of important biological functions. Although some membrane proteins are experimentally well-studied, and even their structures are known, some properties of membrane proteins are intractable by current experimental techniques due to their insolubility and the fact that some functions of membrane proteins require presence of lipid bilayer. With the increasing power of computing, computational studies can sometimes overcome these difficulties and provide insights into the detailed mechanisms of these properties. We are particularly interested in addressing some fundamental questions about ion channels functions regarding their selectivity, gating and protein-lipid interactions using a variety of computational methods, including 3-D structure modeling, and molecular dynamics and Monte Carlo simulations.