CryoHype: Transformer-based hypernetworks for extreme heterogeneity in cryo-EM

ICLR 2026 Conference Submission19733 Authors

Keywords: cryo-EM, hypernetworks, cryo-EM reconstruction, implicit neural representations, neural fields, generalizable neural fields, heterogeneity

TL;DR: Reconstructing extreme compositional heterogeneity in cryo-EM is enabled by hypernetworks

Abstract: Cryo-electron microscopy (cryo-EM) is an indispensable technique for determining the 3D structures of dynamic biomolecular complexes. While typically applied to image a single molecular species, cryo-EM holds great potential for structure determination of many targets simultaneously in a high-throughput fashion. However, existing methods typically focus on modeling conformational heterogeneity within a single or a few structures and are not designed to resolve compositional heterogeneity arising from mixtures of many distinct molecular species. To address this challenge, we propose CryoHype, a transformer-based hypernetwork for cryo-EM reconstruction that dynamically adjusts the weights of an implicit neural representation conditioned on each particle image. CryoHype establishes a new state-of-the-art on the challenging Tomotwin-100 dataset for compositional heterogeneity in CryoBench. We further introduce Sim2Struct-1000, a new synthetic dataset for compositional heterogeneity with 10 times more structures than previous datasets, where CryoHype improves $\mathrm{FSC}_\mathrm{AUC}$ by 67\%. Together, these advances establish transformer hypernetworks as a scalable approach for extreme heterogeneity in cryo-EM reconstruction.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 19733