Fold'EM infers atomic structures directly from cryo-EM particles faster
Fold'EM: Direct atomic structure inference from Cryo-EM particles
Artificial Intelligence
Summary
Determining the shapes of tiny molecules like proteins often involves making a blurry 3D picture from many 2D images, then building a detailed model on top. The authors created Fold'EM, a method that skips the blurry picture step and uses smart protein models to find the exact atomic structure directly from fewer images. It can handle mixed samples with different shapes all at once and works even when the camera angles are unknown. This could make protein structure discovery faster and more efficient.
What this means in practice
- •For biotech development teams: Determine detailed protein structures from fewer cryo-EM images, speeding experimental design and validation.
- •For pharmaceutical researchers: Identify multiple protein shapes directly from mixed cryo-EM data to aid drug targeting of flexible biomolecules.
Authors
Advaith Maddipatla, Märt-Erik Mäeots, Marco Pegoraro, Nikolaus Dräger, Roberto Covino, Sanketh Vedula, Martin Pacesa, Alex M. Bronstein
Abstract
Single-particle cryo-electron microscopy (cryo-EM) has become a widely adopted technique for biomolecular structure determination. The conventional cryo-EM computational pipeline first combines many particle images to reconstruct an electrostatic potential (ESP) map and then fits an atomic model to the recovered map. Density reconstruction has high sample complexity, requiring large numbers of particle images and making structure determination high-cost and low-throughput, particularly for heterogeneous samples. Downstream atomic model building, in turn, becomes increasingly difficult as the resolution of the reconstructed map deteriorates. Protein structure prediction models provide strong sequence-derived priors on atomic structure, and experiment-guided approaches can use these priors to recover structures consistent with experimental measurements. Yet, in cryo-EM, such priors are typically integrated only after density reconstruction during atomic model fitting. We introduce Fold'EM, an inference-time framework that combines priors from protein generative models directly with cryo-EM particle images to determine atomic models from a small number of single particle images, bypassing both intermediate density reconstruction and downstream model building against the reconstructed map. Across synthetic and experimental cryo-EM datasets, Fold'EM recovers accurate atomic structures both with known particle orientations and in an ab-initio setting where orientations are inferred jointly with structure. In heterogeneous datasets, Fold'EM further resolves distinct conformational states from mixed particle populations without separately reconstructing a density map and building an atomic model for each state. We believe these results open new avenues for structure determination in the low-sample regime and for characterizing low-population conformational states directly from cryo-EM particles.