EM Image Processing and Data Analysis

Overall Workflow

In general, workflow looks like this:

1. Set up your account for TACC and 3dem.org.

2. Acquire a stack of serial tSEM images (a complete set of original images with optimized focus, brightness, and contrast; Don't forget to take an image of a calibration grid!)

3. Align the images with AlignEM-SWiFT or Fiji/TrakEM2 on 3dem.org

4. Stack-crop the aligned images, export as .tif files

5. Assign a series code and rename image files

6. Get a spreadsheet ready for series data collection

7. Import aligned images into PyReconstruct

8. Manually fine-tune alignment, if necessary

9. Calibration: x-y (pixel size) with a calibration grid image; z (section thickness) by the cylindrical mitochondria method

10. Choose and trace neuropil elements (and organelles) of interest

11. Data analysis

12. Publication, etc.

See here for what workflow might look like for the KH lab.

Things to do before Reconstructing

Learn about neuropil ultrastructure

!IMPORTANT!  How familiar are you with EM images? Can you identify neuropil, synapses, organelles, etc.?

• Atlas of Ultrastructural Neurocytology and Tutorials

• Peters A, Palay SL, Webster H. (1991) The Fine Structure of the Nervous System (3rd Ed). Oxford University Press, New York. ISBN 0-19-506571-9.

  • For KH lab – we have this book on our file server.

• Electron Micrographs of the Neuropil: Basic Literacy by Patrick Parker

Get your TACC account

!IMPORTANT!  Make sure you have access to the 3dem.org portal at TACC (James Carson is the current contact person).

• Interacting with 3dem.org

Pre-processing of serial section EM images before importing into RECONSTRUCT

• Upload serial section EM images to 3dem.org
  
  

  • Command line instructions (Harris lab specific; restricted access)

• Alignment of serial section EM images using AlignEM-SWiFT

• Assign a series code (Harris lab specific; restricted access)

• Rename aligned serial EM image files with Bulk Rename Utility on Windows PC, or use Bulk Rename software if you are on stampede2 at TACC.
  
  

  • Suggested format: XXXXX_### (XXXXX = series code; ### = section number; 000 = calibration grid image)

Protocols for PyReconstruct

• PyReconstruct User Guide

Protocols for (Legacy) Reconstruct

• Get your Reconstruct here! (Use 32-bit version; 64-bit may be buggy)

• Instructions for new users of Reconstruct: Reconstruct: Intro for Beginners, Reconstruct Basics, User Manual.

• Starting a New Series 

• Section Thickness 

• Calibration Error Correction Strategy
  
  

• Alignment by Correlation

• Adjusting brightness/contrast

• Editing Trace Pallettes

• Importing Traces

• Unbiased 3D stereology

• Grey wall tracing

• Same-dendrite-same-axon analysis

• Dendrites:
  
  

  • Counting or Tracing Mictotubules

  • Tracing Z-Lengths

  • Tracing plinn (linear nearest-neighbor distance between protrusions)

• Spines:
  
  

  • Stamping Dendritic Protrusions

  • Cutting off spines and spine heads: Reconstruct / Blender (with Python add-on described in Bartol et al., 2016) 

• Synapses:
  
  

  • Identifying Symmetric and Asymmetric Synapses

  • Tracing synapses (nascent and active zones)

• Axons:

  • Using Neuropil Tool (CellBlender) to analyze axons

  • Stamping Small Synaptic Vesicles

• Organelles:

  • Ribosome Identification

  • SER and Spine Apparatus Identification

• Exporting quantitative data

• RECONSTRUCT color scheme

Protocols for napari-bootstrapper

Protocols for other analysis tools

• CellBlender: Github | MCell and CellBlender | Tutorial | Neuropil Tools

• NeuroMorph: https://neuromorph.epfl.ch/

• Deprecated Protocols

Data Analysis and Statistics:

• Dendrite Analysis Spreadsheets

• STATISTICA

• Making Box and Whisker Plots in Excel

• Data Transforms

• Statistics for Biologists (Nature collection)

• Estimation for Better Inference in Neuroscience (eNeuro DOI: https://doi.org/10.1523/ENEURO.0205-19.2019 )

• Considerations and recommendations for experimental designs (Journal of Neuroscience)

• Consideration of Sample Size in Neuroscience Studies (Journal of Neuroscience)

• Ask an Expert: Estimation Statistics and Statistical Power (Society for Neuroscience)

• Power Analysis:

  • Introduction to power analysis (UCLA)

  • Sample size determination and power analysis using the G*Power software (J Educ Eval Health Prof, Vol 18, 2021)

  • G*Power (Heinrich-Heine-Universität Düsseldorf)