Linear sampling of large biological molecules using the shadow hybrid Monte Carlo method

Grand scale projects such as protein folding and drug design have put molecular simulation at the forefront of computational research. Even so, inherent limitations in both the models and the methods keep researchers from achieving their goals. Much effort has been expended to improve the traditional simulation methods known as molecular dynamics (MD) and Monte Carlo (MC). Duane and Kennedy[1] were the first to introduce Hybrid Monte Carlo (HMC) which seeks to take advantage of the merits of both MD and MC while forgoing their deficiencies. HMC, however, is not without flaws of its own. [ HMC flaws ] This thesis proposes a new method based on HMC called Shadow Hybrid Monte Carlo (SHMC). SHMC takes advantage of the work of Skeel and Hardy[11] who showed how to inexpensively compute an approximation to the modified Hamiltonian. SHMC has the following properties: [ SHMC properties ].