Viruses are inanimate biomolecular assemblages constructed inside host cells for the tasks of transmission and infection. Despite the high degree of fidelity required to produce functional infectious particles, viruses are adept at developing resistance to natural and vaccine-induced immune responses through escape mutations. This is due to the presence of multiple redundant self-assembly signals, resulting in a fitness landscape with vast regions of favorable sequence space. Viruses can sample extensively from this space to arrive at escape variants, while retaining the capacity to properly assemble. We propose a transformative approach that aims to target large regions of viral fitness landscapes, with the goal of overcoming antiviral drug resistance. In this approach, various viral self-assembly signals will be exploited to induce the mis-assembly of viral components toward non- infectious endpoints. Computationally designed peptides will steer the self-assembly process toward trapped states accessible to large numbers of genetic variants. Peptides are ideal for this task for a number of reasons. Firstly, peptides can target large surface areas of proteins, allowing for binding to targets that lack deep binding pockets. Secondly, peptides can be designed to self-assemble into diverse supramolecular structures, such as fibers, two-dimensional arrays, and liquid condensates. We will leverage the ability of peptides to form these types of structures to induce the formation of non- infectious viral mis-assemblies. Finally, peptides can be optimized for membrane permeability, allowing for the targeting of viral replication inside host cells. Drug-induced mis-assembly of HIV viral capsid has been suggested as one of several possible mechanisms of action of the small-molecule drug PF74. We aim to develop a rationally-guided design approach to enable the wide-spread application of this novel mechanism of anti-viral action. We will target three domains of the HIV Gag protein known to play distinct roles in viral assembly for directed mis-assembly by computationally designed peptides. The resulting peptides will incorporate design elements that direct the trapping of viral components within one-dimensional fibers, two-dimensional arrays, and three-dimensional liquid condensates. Self- assembly will be monitored at the single-molecule level using Interferometric Mass Spectrometry, and at the bulk level with Bio-layer Interferometry and Dynamic Light Scattering, while Electron Microscopy and Fluorescence Microscopy will be used to visualize the induced assemblies. This project paves the way for development of antiviral therapies through mechanisms that pose high barriers to antiviral resistance.