Related Experiment Video
Updated: Jul 2, 2025
09:30
Author Spotlight: Exploring Cellular Processes by Modeling Ligands in Cryo-EM Maps
Published on: July 19, 2024
1.3K
Automated model building and protein identification in cryo-EM maps
Kiarash Jamali1, Lukas Käll2, Rui Zhang3
1MRC Laboratory of Molecular Biology, Cambridge, UK. kjamali@mrc-lmb.cam.ac.uk.
Nature
|February 26, 2024
Summary
ModelAngelo is a new machine-learning tool that automates atomic model building in electron cryo-microscopy (cryo-EM) maps. This AI approach matches human expert quality for proteins and improves nucleotide backbone accuracy, speeding up structural determination.
Area of Science:
- Structural biology
- Computational biology
- Biophysics
Background:
- Atomic model building in electron cryo-microscopy (cryo-EM) maps is complex and requires significant manual effort.
- Current methods demand high expertise and extensive use of 3D graphics software, creating bottlenecks in structure determination.
Purpose of the Study:
- To develop an automated, machine-learning-based approach for atomic model building in cryo-EM maps.
- To improve the efficiency, objectivity, and accuracy of cryo-EM structure determination.
Main Methods:
- ModelAngelo utilizes a graph neural network to integrate cryo-EM map data with protein sequence and structural information.
- It predicts amino acid probabilities for sequence searches using hidden Markov models.
Main Results:
- ModelAngelo builds atomic models for proteins with quality comparable to human experts.
- For nucleotides, ModelAngelo achieves human-level accuracy in backbone construction.
- The tool surpasses human experts in identifying proteins with unknown sequences.
Conclusions:
- ModelAngelo significantly reduces bottlenecks and enhances objectivity in cryo-EM structure determination.
- This automated approach has the potential to accelerate the pace of structural biology research.
