Local computational methods to improve the interpretability and analysis of cryo-EM maps

Area of Science:

• Structural Biology
• Biophysics
• Biochemistry

Background:

• Cryo-electron microscopy (cryo-EM) maps often exhibit heterogeneous B-factors and electron density occupancies.
• Global B-factor sharpening improves contrast but can cause over-sharpening, leading to map distortions and reduced interpretability.
• These distortions hinder accurate ab initio modeling and detailed structural analysis.

Purpose of the Study:

• To develop novel algorithms for enhancing high-resolution features in cryo-EM maps.
• To introduce methods for preventing map distortions caused by over-sharpening.
• To enable the generation of local B-factor and electron density occupancy maps.

Main Methods:

• Utilized the spiral phase transformation as a core component for novel algorithms.
• Developed LocSpiral for enhancing high-resolution features while preventing distortions.
• Introduced LocBSharpen, LocBFactor, and LocOccupancy for local B-factor and occupancy estimations.

Main Results:

• Demonstrated improved enhancement of high-resolution features in cryo-EM maps.
• Successfully prevented map distortions associated with traditional sharpening techniques.
• Generated accurate local B-factor and electron density occupancy maps.
• Applied methods to recent SARS-CoV-2 structures, showing significant improvements.

Conclusions:

• The developed LocSpiral algorithms effectively enhance cryo-EM map interpretability.
• Local B-factor and occupancy maps provide crucial insights for structural analysis.
• These methods offer a significant advancement for processing and analyzing cryo-EM data, particularly for complex biological structures.