Membrane protein structure-based ligand discovery.

Many membrane proteins have been intensively researched and have a lot of known ligands. However, where do these ligands bind and how do they bind are not known for most of the cases. We use docking, together with mutation data, SAR data, selectivity data and potentially many other data, to find binding sites and poses of known ligands of membrane proteins. Once we get the accurate binding poses, we try to understand the mechanism of these molecules.

Discover novel enzymes, metabolites and pathways by pathway docking and genome mining.

The number of protein sequences increase explosively nowadays. However, at least 50% of sequences have their functions unknown, uncertain or incomplete. Traditional bioinformatics approaches use “annotation transfer” to assign known functions to new sequences, and hardly to discover any new functions. In the past, we had used pathway docking and genome mining approach to discover a series of new enzymes, new metabolites and new pathways.

Integrative protein complex modeling.

Many biological processes are done by protein complexes. However, it’s very hard to obtain the structure of protein complexes at high resolution by using signal structural biology methods. Here we are trying develop methods to integrate all structural methods, including X-Ray, EM and NMR data, together with other biophysical or biochemical data such as SXAS, CX-MS, DEER and FRET data, to simulate the structure and dynamics of protein complexes.