Skip to content

Latest commit

 

History

History
87 lines (58 loc) · 4.25 KB

README.md

File metadata and controls

87 lines (58 loc) · 4.25 KB
Raw

FerriTag

Code in this repo is for:

  1. Computer simulations and spatial analysis of FerriTag in Igor Pro
  2. Nanoscale mapping FerriTag distributions using IMOD outputs
  3. Signal-to-noise ratio calculation and 2D Gauss fitting of FerriTag particles
  4. Stereological analysis of particles in EM images

Computer simulations

Igor procedure file Ferritag.ipf featuring:

  • Main program

    • FerriTag() performs the simulation
    • LookAtPDFs() visualise probability density functions
    • MakeFTPlot() compares real data with the model

  • Estimation of the number of observations required to determine the spatial resolution of the method

    • PowerTest()

  • Run the main program many times to test robustness of simulation

    • MedCalc()
    • RunMultiple()

  • Utility functions (called from other functions only)

    • TidyUpPlots()
    • PMMaker()
    • ColorTraces()
    • FlushAllDist()

--

To run the simulation, type LookAtPDFs(7,18), for example.

To generate a 3D figure to show the pose of FerriTag and the measurements taken, run FerriTag(16,6.5,70,100) and append posSect and posWave as scatter in a gizmo. Plasma membrane can be generated using PMMaker().

Note, that the main program requires a wave of observations called FTMeasWave.

To run the simulation many times (settings are hardcoded), type RunMultiple(n) where n is the number of times you'd like to run the program.

To estimate the number of measurements are required, type PowerTest(8,22).

IMOD model analysis

These functions (in IMODModelAnalysis.ipf) will load and analyse FerriTagged particles near clathrin-coated pits visualised by EM.

  1. To start you need to segment your images in IMOD.

  2. Output models as text files using model2point. Use the command model2point -fl -ob -z example_Model_IMOD example.txt for each IMOD model.

  3. These txt files should be in a directory with no other txt files (other filetypes are OK).

  4. Load IMODModelAnalysis.ipf into Igor.

  5. Igor needs scaling information for each image/model to do the analysis. You need to provide this as two waves called:

    1. FileName (textwave) containing the names of each txt file.
    2. PixelSize (numeric) size in nm of a pixel.

    This is important and the analysis will not run without it. We make a csv and load it in using Igor's Load > Data > Load Delimited Text... function.

  6. To run the analysis, select Macros > FerriTag Analysis or run IMODModelAnalysis() from the command line.

  7. Igor will load each model and display it to you with a graphical interface.

  8. Use Cursors A and Cursor B to mark out what you judge to be the start and end of the pit. Drag the Cursors from the little box below the graph.

  9. Click continue and the next one will display until you have loaded all models from your directory.

  10. Igor will then display several graphs of the analysis.

There are two layouts which can be saved in a variety of formats.

2D Gaussian fitting and SNR calculation

In the file FTAnalysisSNR.ipf, you will find tools to fit 2D Gaussian functions to FerriTag particles. The location of the peaks is then used to calculate the SNR. The function LoadTIFFFilesForAnalysis() works on a directory of TIFFs, but it requires several waves loaded into the experiment to do this properly. See comments for details.

Stereology

The locations of particles in images are recorded in ImageJ and saved as a csv. To do this a images are stripped of metadata and the filenames encoded (using blind_analysis. ijm) so that the person locating particles is blind to the conditions of the experiment.

A segmented version of each EM image is also required (in a directory called Masks). See comments for further details.

To run Stereology Workflow the following files are needed:

  1. Images and a sub directory of Masks
  2. csv with locations of particle xy coords
  3. log file from blind_analysis for unblinding the images
  4. csv with pixelsize information
  5. csv with a list of conditions in each image (encoded with an integer).