Classification of structural heterogeneity by maximum-likelihood methods.

Area of Science:

• Structural biology
• Biophysics
• Computational imaging

Background:

• Maximum-likelihood (ML) image processing is now computationally feasible for cryo-electron microscopy (cryo-EM).
• These ML methods are highly effective for classifying structurally heterogeneous single-particle data.
• Numerous studies utilize these algorithms for analyzing macromolecular complexes exhibiting significant structural diversity.

Purpose of the Study:

• To share practical experiences gained from applying novel ML approaches in cryo-EM.
• To provide insights into data preparation and classification strategies for heterogeneous samples.
• To assist microscopists in applying ML methods to their research.

Main Methods:

• Application of computationally feasible maximum-likelihood (ML) image processing algorithms.
• Classification of single-particle cryo-electron microscopy data.
• Two- and three-dimensional classification techniques.
• Considerations for high-performance computing.

Main Results:

• ML methods successfully classify structurally heterogeneous single-particle data in cryo-EM.
• These approaches are effective for analyzing complexes with nonstoichiometric formation and large conformational changes.
• Practical experience has been gained in applying these novel computational techniques.

Conclusions:

• ML methods represent a significant advancement for analyzing structural variability in macromolecular complexes using cryo-EM.
• This chapter offers practical guidance for researchers implementing ML-based classification.
• Successful application requires careful data preparation and understanding of computational aspects.