a b s t r a c t
Simulations of virus-like particles needed for computer-aided vaccine design highlight the need for new
algorithms that accelerate molecular dynamics. Such simulations via conventional molecular dynamics
present a practical challenge due to the millions of atoms involved and the long timescales of the
phenomena of interest. These phenomena include structural transitions, self-assembly, and interaction
with a cell surface. A promising approach for addressing this challenge is multiscale factorization. The
approach is distinct from coarse-graining techniques in that it (1) avoids the need for conjecturing phenomenological
governing equations for coarse-grained variables, (2) provides simulations with atomic
resolution, (3) captures the cross-talk between disturbances at the atomic and the whole virus-like particle
scale, and (4) achieves significant speedup over molecular dynamics. A brief review of multiscale
factorization method is provided, as is a prospective on its development.
a b s t r a c tSimulations of virus-like particles needed for computer-aided vaccine design highlight the need for newalgorithms that accelerate molecular dynamics. Such simulations via conventional molecular dynamicspresent a practical challenge due to the millions of atoms involved and the long timescales of thephenomena of interest. These phenomena include structural transitions, self-assembly, and interactionwith a cell surface. A promising approach for addressing this challenge is multiscale factorization. Theapproach is distinct from coarse-graining techniques in that it (1) avoids the need for conjecturing phenomenologicalgoverning equations for coarse-grained variables, (2) provides simulations with atomicresolution, (3) captures the cross-talk between disturbances at the atomic and the whole virus-like particlescale, and (4) achieves significant speedup over molecular dynamics. A brief review of multiscalefactorization method is provided, as is a prospective on its development.
การแปล กรุณารอสักครู่..
