Recognizing the fact that many novel drugs failed to pass clinical evaluation is mainly due to their poor ADME/T properties, the recent focus of our lab has been developing an integrated computational methodology that can help predict the ADME/T properties of new candidate compounds in the early phases of drug development. The first part of this integrated methodology includes identification of substrates or inhibitors of metabolic enzymes, e.g., cytochrome P450 and the efflux pump, e.g., P-glycoprotein, which will facilitate homology modeling, docking, and molecular dynamics simulations, and binding free energy calculations. The second part will include QSAR prediction of intestinal permeability, blood brain barrier, among other important ADME/T properties based on passive transport mechanism.

In the past, we have proposed the "relaxed complex" scheme that can accommodate receptor flexibility for structure-based drug design, and it has been demonstrated that this computational approach can reach the same accuracy as the "SAR by NMR" method for lead optimization. Along the same line we will continue to explore the significance of protein flexibility for molecular recognition.

Besides the development of new computational methodology, we have also been developing new antitumor agents based on virtual screening of large chemical databases. The current approach is to design a potent inhibitor to block the function of farnesyltransferase, which plays a very important role for the post-translational modification of the Ras protein.

Continued with our previous work, we are also investigating the interactions of toxic peptides with biological membranes. We have applied the Atomic Force Microscopy simulations to unravel the interaction of a single peptide and a peptide channel with a lipid bilayer. This could provide us with invaluable dynamical insights from such non-equilibrium processes. We are also investigating the molecular mechanism of selected targeting to cancer cells of some toxic peptides.

In addition, we are also investigating the activation and gating mechanisms of ion channels, as well as the selection permeation of ion channels. The methodologies employed are molecular dynamics and Brownian dynamics simulations. We could also evaluate the association rate constants of membrane proteins to a vesicle under various conditions based on massive Brownian simulation.

We welcome people with different background, e.g., physics, chemistry, mathematics, computer sciences, pharmacy, and so on, to join our lab for these multidisciplinary researches.