SUMMARY The field of structural biology is experiencing transformative change. New experimental and computational tools are revealing a remarkable catalogue of structures with unprecedented speed. In addition to the thousands of determined structures and the millions of new structure predictions, novel protein designs are emerging in a flurry facilitated by a host of new machine learning approaches. In the face of all this, important gaps remain that limit our fundamental understanding of protein structure and function, and our ability to rapidly visualize atomic molecular structures. These gaps persist where conventional approaches have been traditionally stalled, for example in the characterization of large, multimeric protein complexes, unknown chemical entities and dynamic assemblies. The amyloid fold is a notable example of such a challenging area, given its long lack of atomic resolution structures and viable therapeutics, as well as its emerging yet enigmatic role in benign physiological processes such as phase separation by low complexity sequences. The Rodriguez lab has been at the forefront of efforts to uncover complex amyloid structures with remarkable detail, facilitated in part by an ESI MIRA (R35 GM128867) award. That award also funded critical developments that brought to life new methods for structure determination via MicroED. However, despite the promise of these methods, the single isolated structures they have revealed in remarkable detail have lacked context and represent only a small part of the structural path from single molecule to fibril. The current proposal, a renewal of R35 GM128867, builds on that successful five year effort to characterize toxic and infectious assemblies at the atomic level and envisions three directions for an improved structural understanding of protein assemblies: We now propose to 1. Further the development of new tools for more accurate and rapid determination of molecular and cellular structures. 2. Advance our fundamental understanding of highly pleiomorphic molecules, particularly those that are prone to forming large- scale multimeric assemblies through amyloid-like interactions, in functional or pathogenic roles. 3. Chart sequence-function relationships that allow proteins to give rise to and/or facilitate pathogenic traits through allosteric or epistatic networks of residue level changes, particularly in acutely lethal agents. These directions continue to elaborate rigorous approaches to molecular structure determination, while leveraging the compendium of emerging tools for structure prediction and discovery. Hybrid approaches allow us to take on the more ambitious challenges of charting the function of pathogenic and functional amyloids in situ, and to chart vast functional sequence landscapes in dynamic multiprotein assemblies. Collectively, the efforts bring us closer to the ideal of witnessing atomic structures in context and uncovering the basis for molecular processes that underpi...