The dream of any biologist is to have an "Asimovian nanosubmarine" to allow the observations of molecular scale events first hand. While such a device does not currently exist, biologists and chemists can instead employ multiscale modeling techniques to investigate molecular behaviors. By integrating kinetic parameters, structural data, and computation, it is possible to animate cellular scenes. While the construction of such a model may seem simple, the task requires the assimilation of parameters from literature, and the generation of computable representations of cellular structures. Where parameters are missing, computational approaches such as molecular dynamics can be used to supplement by predicting values of interest. In this dissertation, I describe my efforts in computing important physical properties such as passive membrane permeability of drug molecules, and binding/unbinding kinetics using molecular dynamics and a variety of enhanced sampling strategies. Beyond these transport properties, the conformational dynamics of proteins is critical to understand problems such as drug efficacy and protein function. I present a review of emerging computational methods to rationally design allosteric drugs, and a study using Markov state modeling theory to represent the differential dynamics of C-C chemokine receptor type 2 with and without drugs bound. To address the computer representation of cellular structures, I present an implementation of a general simplicial complex (mesh triangulations) data structure suitable for mesh generation applications. Collectively the work presented in this dissertation address the barriers hindering the development of physical models of whole cells on multiple fronts.