Questions: Single-Particle Analysis

Question 1

Why does single-particle cryo-EM require averaging tens of thousands to millions of individual particle images?

• A. Because each particle image is unique and cannot be compared to others
• B. Because each individual particle image has extremely low signal-to-noise ratio (the electron dose must be kept low to prevent radiation damage), and averaging many images of particles in the same orientation builds up the signal while averaging out the noise
• C. Because the microscope can only image one particle at a time
• D. Because statistical significance requires large sample sizes for any measurement

Answer: (B)

Biological specimens are radiation-sensitive — high electron doses destroy the structure. Each particle image is therefore taken with very few electrons (~20-40 electrons per Angstrom squared), resulting in extremely noisy images where the protein signal is barely distinguishable from noise. Averaging N images in the same orientation improves the signal-to-noise ratio by a factor of sqrt(N). To achieve near-atomic resolution, the SNR must be improved by a factor of ~100-1000, requiring tens of thousands to millions of particle images. This is why cryo-EM is fundamentally a statistical method — the resolution depends on having enough particles to average.

Question 2

3D classification in single-particle analysis can only separate particles by their orientation, not by their conformation.

• T. True
• F. False

Answer: False

3D classification is one of cryo-EM's most powerful capabilities — it can sort particles into groups that differ in both orientation AND conformation. If a sample contains molecules in multiple functional states (e.g., a ribosome in pre- and post-translocation states, or a channel in open and closed conformations), 3D classification assigns each particle to the conformational class it best matches. The result is separate 3D reconstructions for each conformational state, revealing the structural basis of functional dynamics. This is a major advantage over X-ray crystallography, where the crystal lattice typically selects for a single conformation, and heterogeneity smears out the electron density.

Question 3

Explain the problem of preferred orientation in cryo-EM and why it limits the achievable resolution.

Model answer: Single-particle reconstruction requires views of the molecule from many different orientations to reconstruct the full 3D structure — analogous to how CT scanning requires X-ray projections from many angles. If particles preferentially adsorb to the ice-air interface in one orientation (preferred orientation), the dataset lacks views from certain directions. The missing orientations create an 'angular gap' in Fourier space, producing an anisotropic reconstruction that is well-resolved in some directions but smeared in others (like a CT scan with missing projections). Severe preferred orientation can prevent structure determination entirely. Solutions include tilting the grid during data collection, using different grid surfaces, adding surfactants, or engineering the protein with asymmetric features.

Preferred orientation is one of the most common practical problems in cryo-EM. Many proteins have a flat face that preferentially sits on the ice surface, producing mostly 'top views' with few 'side views.' The problem is exacerbated on thin ice films, where both surfaces constrain particle orientation.