CryoDECO: Deconstructing Extreme Compositional and Conformational Heterogeneity in Cryo-EM via Foundation Model Priors
Published in LangTaoSha Preprint Server, 2026
Abstract
CryoDECO addresses extreme compositional and conformational heterogeneity in cryo-EM by using foundation-model priors to guide heterogeneous reconstruction. The approach reduces reliance on random ab initio initialization and helps separate compositional classification from three-dimensional reconstruction. By using learned particle representations as priors, CryoDECO aims to deconstruct complex native-like mixtures and support panoramic structural analysis of heterogeneous molecular assemblies.
Recommended citation: Yang Yan, Yanwanyu Xi, Shiqi Fan, Yifei Wang, Ziyun Tang, Fajie Yuan, Huaizong Shen. (2026). "CryoDECO: Deconstructing Extreme Compositional and Conformational Heterogeneity in Cryo-EM via Foundation Model Priors." LangTaoSha Preprint Server. DOI: 10.65215/LTSpreprints.2025.12.30.000075.
Publisher page