Molecular simulation has become an important tool in modern computational chemistry and biochemistry. Nevertheless
accuracy and efficiency of the approach still need further improvement to achieve the goal of robust and predictive simulation, particularly for large and complex biomolecular systems. The accuracy issue arises from the intrinsic limitations of classical models
that have to be used to approximate the quantum molecular processes. The efficiency issue is a direct consequence of the high dimensionality of biomolecular systems: sophisticated molecular machines are complexes of thousands to millions of atoms. What further
complicates the picture is the need to realistically model the interactions between biomolecules and their surrounding water molecules. In this talk, I will review our developments of biomolecular simulation models and methods and their applications to interesting
biomedical systems. For developments, I will go over our recent works for more accurate biomolecular force fields for explicit or implicit consideration of electronic polarization. I will also highlight our developments for efficient modeling solvation-mediated
energetics and dynamics, covering both polar and nonpolar interactions, crucial in receptor-ligand modeling. For applications, I will highlight several of our collaborative efforts addressing interesting biological problems, in both cancer biology and biofuel
synthesis.