CryoTriplane: triplane-based neural radiance field representation for high-resolution cryo-EM 3D homogeneous reconstruction.
4 affiliations
- 1Frontiers Science Center for Nonlinear Expectations (Ministry of Education), Research Center for Mathematics and Interdisciplinary Sciences, Qilu Hospital (Qingdao), Cheeloo College of Medicine, Shandong University, Qingdao 266237, China.
- 2College of Medical Information and Engineering, Ningxia Medical University, Yinchuan 750004, China.
- 3School of Computer Science and Technology, Shandong University, Qingdao 266237, China.
- 4School of Medical Technology, Beijing Institute of Technology, Beijing 100190, China.
Abstract
Motivation
Determining the three-dimensional (3D) structure of proteins is a cornerstone of structural biology, providing essential insights into their underlying biological mechanisms and functional roles. Cryogenic electron microscopy (cryo-EM) single-particle analysis allows researchers to investigate biomolecules at near-atomic resolutions, unraveling the structure of biomolecular complexes. However, reconstructing high-fidelity 3D volumes from extremely noisy projections, particularly under the interference of structural heterogeneity, remains a challenging task. Traditional Fourier-domain methods often suffer from frequency-dependent signal attenuation and uneven information distribution. While recent deep-learning-based frameworks attempt to predict high-frequency components directly in frequency space, they are highly susceptible to noise and often introduce artifacts. State-of-the-art (SOTA) real-space method circumvents Fourier-domain defects, yet remains constrained by the inherent spectral bias of multi-layer perceptrons, as fixed positional encoding fails to sufficiently capture intricate high-frequency details.Results
We propose cryoTriplane, a real-space 3D homogeneous reconstruction network that utilizes a hybrid spatial representation to adaptively capture fine structural details. By combining fixed positional encoding for noise-robust structural initialization with structure-aware positional encoding for high-frequency refinement, our method achieves 3D reconstructions that are simultaneously globally consistent and locally fine-grained. Comprehensive qualitative and quantitative evaluations across five simulated and four real-world datasets demonstrate that cryoTriplane significantly outperforms existing methods in resolving fine structural features. Furthermore, validation through downstream atomic modeling underscores cryoTriplane's ability to recover high-quality density maps. Specifically, our method achieves sequence coverage improvements of 4.6% and 5.5% on the EMPIAR-10005 and EMPIAR-10492 datasets over the SOTA approach, respectively.Availability and implementation
Our code is available at https://github.com/chengyr1999/CryoTriplane. The real datasets used in this study are publicly available from EMPIAR.
Related
- Closing the multichannel gap through computational reconstruction of interaction in super-resolution microscopy
- Characterization of spatial homogeneous regions in tissues with concordex
- A Fluorescent Aptamer Sensor Based on Cu-TCPP/rGO for Homogeneous Screening of G-Quadruplex Ligands
- Immobilization of Microbial Transglutaminase for the Production of Homogeneous Immunotherapeutic Antibody-Drug Conjugates.
- Proline iminopeptidase: a novel in vivo strategy to produce proteins with homogeneous and authentic N-termini in E. coli.
BibTeX
@article{cheng2026cryotriplane,
title = {{CryoTriplane: triplane-based neural radiance field representation for high-resolution cryo-EM 3D homogeneous reconstruction.}},
author = {Yiran Cheng and Bintao He and Qiong Zeng and Fa Zhang and Renmin Han},
journal = {Bioinformatics (Oxford, England)},
year = {2026},
month = oct,
doi = {10.1093/bioinformatics/btag754},
url = {https://europepmc.org/article/MED/42870393},
note = {CC BY},
}
RIS
TY - JOUR
TI - CryoTriplane: triplane-based neural radiance field representation for high-resolution cryo-EM 3D homogeneous reconstruction.
AU - Yiran Cheng
AU - Bintao He
AU - Qiong Zeng
AU - Fa Zhang
AU - Renmin Han
PY - 2026
DA - 2026/10/01
JO - Bioinformatics (Oxford, England)
DO - 10.1093/bioinformatics/btag754
UR - https://europepmc.org/article/MED/42870393
AB - Motivation Determining the three-dimensional (3D) structure of proteins is a cornerstone of structural biology, providing essential insights into their underlying biological mechanisms and functional roles. Cryogenic electron microscopy (cryo-EM) single-particle analysis allows researchers to investigate biomolecules at near-atomic resolutions, unraveling the structure of biomolecular complexes. However, reconstructing high-fidelity 3D volumes from extremely noisy projections, particularly under the interference of structural heterogeneity, remains a challenging task. Traditional Fourier-domain methods often suffer from frequency-dependent signal attenuation and uneven information distribution. While recent deep-learning-based frameworks attempt to predict high-frequency components directly in frequency space, they are highly susceptible to noise and often introduce artifacts. State-of-the-art (SOTA) real-space method circumvents Fourier-domain defects, yet remains constrained by the inherent spectral bias of multi-layer perceptrons, as fixed positional encoding fails to sufficiently capture intricate high-frequency details.Results We propose cryoTriplane, a real-space 3D homogeneous reconstruction network that utilizes a hybrid spatial representation to adaptively capture fine structural details. By combining fixed positional encoding for noise-robust structural initialization with structure-aware positional encoding for high-frequency refinement, our method achieves 3D reconstructions that are simultaneously globally consistent and locally fine-grained. Comprehensive qualitative and quantitative evaluations across five simulated and four real-world datasets demonstrate that cryoTriplane significantly outperforms existing methods in resolving fine structural features. Furthermore, validation through downstream atomic modeling underscores cryoTriplane's ability to recover high-quality density maps. Specifically, our method achieves sequence coverage improvements of 4.6% and 5.5% on the EMPIAR-10005 and EMPIAR-10492 datasets over the SOTA approach, respectively.Availability and implementation Our code is available at https://github.com/chengyr1999/CryoTriplane. The real datasets used in this study are publicly available from EMPIAR.
SN - 1367-4803
SN - 1367-4811
ER -