Deriving topology and sequence alignment for the helix skeleton in low-resolution protein density maps

Area of Science:

• Structural Biology
• Computational Biology
• Biophysics

Background:

• Cryoelectron microscopy (cryoEM) determines 3D protein structures at resolutions visualizing secondary structures like alpha-helices and beta-sheets.
• Predicting the complete protein backbone from cryoEM density maps remains a challenge.

Purpose of the Study:

• To develop a computational method for predicting the topology and sequence alignment of alpha-helices from cryoEM density maps.
• To improve protein structure prediction by leveraging secondary structure information.

Main Methods:

• Combined geometrical analysis of skeleton helices with Rosetta ab initio structure prediction.
• Developed a consensus approach for topology and sequence alignment prediction.
• Utilized skeleton helix topology for clustering decoy structures.

Main Results:

• Successfully predicted correct helix topology for the majority of helices in 45 out of 60 tested proteins.
• Achieved sequence alignment offsets within +/-2 amino acids for most assigned helices.
• Demonstrated that topology clustering is superior to overlap clustering for refining decoy structures, especially with small decoy sets.

Conclusions:

• The developed method effectively predicts alpha-helix topology and sequence alignment from cryoEM data.
• This approach aids in reconstructing protein backbones from experimental density maps.
• Topology-based clustering offers an improved strategy for analyzing protein structure prediction decoys.