Beyond consensus refinements: Advanced cryo-EM data processing with CryoSPARC

Workshop Description:

One of the main goals of single-particle analysis is the production of a consensus map, in which all particles are aligned to a single reference volume. For some samples, a single map is all that’s needed. However, many samples require additional analysis after the first good map is produced. Perhaps the sample is pseudosymmetric and each asymmetric unit needs to be treated separately, or there is some flexibility which is blurring part of the map, or there are actually two or more distinct particle populations which must be separated.

These questions cannot be tackled with consensus refinement alone — they must be addressed with processing jobs which consider subpopulations of the particle stack and which are able to handle particle heterogeneity. Which job is appropriate and how that job should be run are both sample dependent. Attendees of this two-day workshop will learn how to recognize when these heterogeneous subpopulations may exist as well as the theory and practice of processing jobs which help investigate them.

This workshop will be taught by experts from the Structura Biotechnology team, which builds CryoSPARC. You will have an opportunity to interact with the instructors and fellow attendees and advance your understanding of and ability to perform single-particle data processing.

Workshop Format:

This two-day workshop is centered around an interactive lecture format. Attendees will:

• set up and run CryoSPARC jobs,
• investigate 3D maps and diagnostic plots in detail,
• explore pre-run jobs,
• discuss the reasoning behind which job types are used and why, and
• analyze the effects of modifying specific key parameters.

Additionally, time is reserved for detailed group discussion of a limited number of case studies brought forward by attendees.

Background Concepts:

Applicants should ideally already have a sound understanding of the following concepts in cryo-EM image analysis in order to gain the most from this workshop:

• Fourier analysis
  • Spatial signals can be represented directly (real space) or as a sum of simple waves (Fourier space).
  • Waves with a lower frequency are closer to the origin; waves with a higher frequency are further from the origin.
  • The angular position of a wave in Fourier space indicates the direction that wave travels in real space.
• Refinement fundamentals
  • A particle image is related to a reference by a set of rotations and translations, together called a pose.
  • New references are generated from particle images and poses in a process called backprojection (or, equivalently, Fourier insertion).
• Fourier Shell Correlation (FSC)
  • The FSC measures the correlation between two volumes. At each frequency shell, the correlation between those two maps (ranging from zero to one) is measured. This frequency-correlation relationship forms the FSC curve.
• Contrast Transfer Function (CTF)
  • Cryo-EM images are collected with some amount of defocus. This defocus creates phase contrast in the images, but also creates aberrations which must be computationally corrected.