Self-assembly is a process whereby molecules organize into structures with hierarchical order and complexity, often leading to functional materials. Biomolecules such as peptides, lipids and DNA are frequently involved in self-assembly, and this leads to materials of interest for a wide variety of applications in biomedicine, photonics, electronics, mechanics, etc. The diversity of structures and functions that can be produced provides motivation for developing theoretical models that can be used for a molecular-level description of these materials. Here we overview recently developed computational methods for modeling the self-assembly of peptide amphiphiles (PA) into supramolecular structures that form cylindrical nanoscale fibers using molecular-dynamics simulations. Both all-atom and coarse-grained force field methods are described, and we emphasize how these calculations contribute insight into fiber structure, including the importance of beta-sheet formation. We show that the temperature at which self-assembly takes place affects the conformations of PA chains, resulting in cylindrical nanofibers with higher beta-sheet content as temperature increases. We also present a new high-density PA model that shows long network formation of beta-sheets along the long axis of the fiber, a result that correlates with some experiments. The beta-sheet network is mostly helical in nature which helps to maintain strong interactions between the PAs both radially and longitudinally. Copyright (C) EPLA, 2017