Bayesian algorithms for recovering structure from single-particle diffraction snapshots of unknown orientation: a...

Area of Science:

• Structural biology
• X-ray physics
• Biophysics

Background:

• X-ray free-electron lasers (XFELs) enable single-particle imaging before destruction.
• Determining particle orientation from low-photon diffraction data is crucial.
• Previous studies presented distinct methods for orientation determination.

Purpose of the Study:

• To unify apparently different orientation determination approaches for single-particle X-ray imaging.
• To identify the main factors limiting the performance of these methods.
• To advance structural determination of delicate biological macromolecules and viruses.

Main Methods:

• Analysis of two distinct computational approaches for particle orientation retrieval.
• Comparison of algorithms under low signal-to-noise conditions (~10(-2) photons per pixel).
• Identification of core mathematical and computational principles underlying both methods.

Main Results:

• The two presented approaches are demonstrated to be different implementations of the same core algorithm.
• Key factors limiting orientation determination accuracy and success rate were identified.
• The findings provide a unified framework for understanding and improving single-particle orientation methods.

Conclusions:

• A unified understanding of single-particle orientation determination methods is established.
• Performance limitations are clearly defined, guiding future algorithm development.
• This work facilitates more robust structural analysis of individual biomolecules and viruses using XFELs.