The Electrostatic Effect of Infinite Solvent

• Main Authors: Yang, Pei-Kun, 楊佩琨
• Other Authors: Liaw, Shwu-Huey
• Format: Others
• Language: zh-TW
• Published: 2002
• Online Access: http://ndltd.ncl.edu.tw/handle/02641583864399092984

博士 === 國立臺灣大學 === 分子醫學研究所 === 91 === Water plays critical roles in organisms. For example, soluble, globular proteins are stable in water, but may be denatured in other solvents; in many metabolic reactions, a water molecule acts as a nucleophile. To understand the dynamic behavior of macromolecules in atomic detail, the effect from water molecules must be included in molecular dynamics (MD) simulations. One way to achieve this is to treat water molecules far from a solute implicitly, and to estimate their electrostatic contributions by solving the Poisson-Boltzman equation using finite-difference methods. In Chapter one, we derive an analytical solution of the Poisson-Boltzman equation with rectangular boundary conditions (referred to as the Rectangular Image Charge or RIC method), which was subsequently verified by finite-difference solution of the Poisson-Boltzman equation. The van der Waals contributions from implicit water molecules outside the simulation box were approximated by a polynomial function. The implicit water electrostatic and vdW interaction forces and energies were incorporated into the CHARMM program, and cation-water oxygen radial distribution functions and absolute free energies of ions were computed. The absolute free energy of an ion was verified to be independent of its location in the cubic box and the size of box, indicating that the RIC method can extend the finite system to an infinite solvent system. In the chapters two and three, we prove that by placing a charged/uncharged solute in an infinite water system, the electrostatic potential of the solute, and thus the charging free energy are divergent. Consequently, we developed a strategy using a macroscopic water droplet to mimic the real system in computing the electrostatic potential and charging free energy. We showed that the continuum model should not strictly be used in a system with only one explicit/implicit water boundary as in conventional calculations. Instead, it should be used in a system with two explicit/implicit water boundaries, and include the electrostatic potential difference between the two boundaries in continuum dielectric methods.