CryoACE: An Atom-centric Framework for Accurate and Automated Model Building in Cryo-EM

2026-06-30

AI Summary

This study addresses the challenge of simultaneously ensuring physicochemical plausibility and capturing conformational heterogeneity in automated protein modeling from cryo-electron microscopy (cryo-EM) density maps. To this end, the authors propose CryoACE, an end-to-end framework that introduces an atom-centered reconstruction paradigm. By directly sampling density features at atomic coordinates and iteratively refining structures, CryoACE eliminates conventional voxel-based convolutions, enabling efficient multimodal integration. The method further incorporates a training-free, local-resolution prior to guide conformational sampling, effectively resolving dynamic ambiguities. CryoACE substantially outperforms existing approaches on static benchmarks and, for the first time, reveals atomic-level conformational dynamics in real datasets such as EMPIAR-10345 without requiring a pre-built static model.

Abstract

Protein automodeling from cryo-EM density maps faces unique challenges in enforcing physicochemical validity and managing conformational heterogeneity. Current solvers are often limited to static predictions or require computationally intensive heuristic searches. We present CryoACE, an end-to-end framework that reconstructs precise atomic graphs for both homogeneous and heterogeneous structures. Our method features two key innovations: an atom-centric reconstruction paradigm, where density features are sampled directly at atomic coordinates and iteratively recycled to refine structures, replacing expensive voxel convolutions for efficient multimodal fusion; and a training-free guidance mechanism that leverages predicted local resolution priors to resolve dynamic ambiguity. Validated on a newly constructed high-quality dataset, CryoACE significantly outperforms existing baselines on static benchmarks and, for the first time, unveils atomic-level dynamic conformations on complex real-world datasets like EMPIAR-10345 without relying on pre-built static structures.

Problem

• cryo-EM
• protein automodeling
• conformational heterogeneity
• physicochemical validity
• atomic model building

Innovation

• atom-centric reconstruction
• training-free guidance
• cryo-EM model building
• conformational heterogeneity
• local resolution prior