My research is broadly concerned with the development of computational methods and their applications to disease-related biomolecules. In close collaboration with experimentalists, I use a number of simulation and modeling techniques, including molecular dynamics, virtual screening, free energy calculations, and enhanced sampling methods, to study various biological systems. Currently, my work includes the following three directions:

M. tuberculosis (Mtb) is the primary causative pathogen of tuberculosis, an infectious disease that kills nearly two million people every year. My work is aimed at identifying novel inhibitors against key enzymes required for the synthesis of the Mtb cell wall.

In collaboration with the group of Prof. Michael R. McNeil, we have studied two enzyme targets, RmlC and RmlD, which are involved in the synthesis of L-rhamnose, a deoxy sugar that helps to link the arabinogalactan and peptidoglycan layers in the Mtb cell wall. Our chemical and virtual screening has led to the discovery of low micromolar inhibitors to both enzymes.

Since the natural resistance of Mtb to antibiotics is at least partly due to the permeation barrier of its cell wall, another aim of my research is to develop a computational model that predicts the pathway and permeability of drug molecules in this cellular envelope of the mycobacterium.

I. Computer-aided drug design targeting M. tuberculosis cell wall synthesis

Virtual screening against M. tuberculosis enzyme RmlD

II. Development and application of enhanced sampling methods

Molecular dynamics (MD) is a widely used computational method in the study of protein folding, transmembrane permeation, ligand-receptor binding, and various other biological or chemical problems. Apart from conventional MD, various enhanced sampling methods are used to improve sampling of the conformational space. Among them is accelerated molecular dynamics (aMD), a recently developed method that achieves enhanced sampling by reducing energy barriers separating different states.

Together with my former colleagues at the University of Illinois, Urbana-Champaign, we implemented the aMD method into the simulation program NAMD (release 2.8). This implementation allows aMD simulations of large (>100,000 atoms) biomolecular systems using thousands of processors. Following the initial testing on a peptide, we have applied the method to enhance the diffusion and mixing of lipid bilayers. The results provide the basis for simulations of complex lipid mixtures that best resemble a realistic biological membrane.

The accelerated molecular dynamics method

III. Selectivity of Antimicrobial peptides

Antimicrobial peptides (AMPs) are the first line of defense against infection in most living organisms. Many of these small, cationic, and amphipathic peptides target the negatively charged bacterial membranes, where they form pores once the peptide/lipid ratio reaches a certain threshold.

In collaboration with the group of Prof. Judy E. Kim, my research in this area focuses on elucidating the molecular basis of AMP selectivity for mammalian and bacterial membranes. Using the synthetic AMP CM15 as a model system, we have investigated its interactions with different lipid bilayers via a number of methods, such as molecular dynamics simulations, fluorescence and Raman spectroscopy, as well as circular dichroism. Ongoing and future work includes the joined theoretical and experimental study of several other AMPs.

Antimicrobial peptides interacting with a lipid bilayer

During my PhD, I studied the selectivity, gating, and transport of membrane proteins, such as the water channel aquaporin and the transporter ADP/ATP carrier. Some of my graduate work can be found here (membrane channels and transporters).