Related Experiment Video
Updated: Jul 10, 2025
13:43
A Robust Single-Particle Cryo-Electron Microscopy cryo-EM Processing Workflow with cryoSPARC, RELION, and Scipion
Published on: January 31, 2022
13.3K
Integrating AlphaFold and deep learning for atomistic interpretation of cryo-EM maps
Xin Dai1, Longlong Wu2, Shinjae Yoo1
1Computational Science Initiative, Brookhaven National Laboratory, Upton, NY, USA.
Briefings in Bioinformatics
|November 20, 2023
Summary
This study introduces CryoFEM, a deep learning tool that enhances cryo-electron microscopy (cryo-EM) maps. Combined with AlphaFold, it significantly improves the accuracy of atomic models for biological molecules.
Area of Science:
- Structural Biology
- Biophysics
- Computational Biology
Background:
- Accurate 3D atomic models are crucial for interpreting cryo-electron microscopy (cryo-EM) maps.
- AlphaFold provides initial 3D coordinates, but manual model building is often required due to inaccuracies and conformational heterogeneity.
Purpose of the Study:
- To develop an automated workflow for building and fitting atomic models into cryo-EM maps.
- To enhance the accuracy of cryo-EM map interpretation using deep learning and protein structure prediction.
Main Methods:
- A novel workflow integrating CryoFEM (Cryo-EM Feature Enhancement Model), a deep learning map enhancement tool, with AlphaFold.
- Benchmarking the workflow on 36 cryo-EM maps to assess performance against ground truth models.
Main Results:
- CryoFEM demonstrated state-of-the-art performance in optimizing Fourier Shell Correlations between cryo-EM maps and ground truth models.
- The integrated workflow significantly improved model accuracy for 17 datasets with initially less accurate AlphaFold predictions.
Conclusions:
- The integration of deep learning-based image enhancement (CryoFEM) and AlphaFold offers a promising approach for automated model building and fitting.
- This advancement facilitates more efficient and accurate atomistic interpretation of cryo-EM data.
