A pocket-centric framework for selective targeting of amyloid fibril polymorphs

The rapid expansion of high-resolution cryo-EM structures of amyloid fibrils has not yet translated into the rational design of selective or specific ligands of protein aggregates involved in Alzheimers and Parkinsons diseases. This persistent limitation suggests that the obstacle lies into a certain degree of communality within the organization of fibrillar polymorphs surfaces available for small molecule binding. Here, we present a systematic and global analysis of binding pockets across 97 cryo-EM structures of amyloid-{beta}, tau, and -synuclein protein fibrillar polymorphs. Using a unified pocket similarity index and minimum spanning tree representations, we construct global and protein-specific "pocketomes" that reveal how surface cavities are distributed across different amyloid-forming proteins and the fibrillar polymorphs they form. We show that most detectable pockets are shared across multiple fibrillar folds and, in many cases, across different amyloid-forming proteins, providing a structural explanation for the widespread lack of ligand selectivity. Conversely, a limited subset of pockets forms isolated clusters associated with specific proteins or polymorphs, delineating the rare structural conditions under which selective or specific ligand design is feasible. Together, these results reframe amyloid targeting as a problem of constrained pocket diversity within the amyloid polymorphs landscape, and provide a conceptual framework to guide both the design of future ligands and the strategic avoidance of intrinsically non-discriminatory binding sites. Significance StatementDespite major advances in cryo-EM structure determination of amyloid fibrils, the development of selective ligands for such assemblies remains largely unsuccessful. By systematically comparing surface binding pockets across nearly one hundred amyloid-{beta}, tau, and -synuclein fibrillar structures, we show that this failure is rooted in the global similarity of fibril interaction pockets across amyloid-forming proteins and the various resulting polymorphs. Our analysis reveals that only a small fraction of pockets displays the structural isolation required for protein-selective or polymorph-specific targeting, while most are intrinsically prone to off-target binding. This work provides a structural rationale for decades of limited ligand specificity and establishes a pocket-centric framework to guide realistic design strategies for amyloid imaging and therapeutics.