Skip to content
This repository has been archived by the owner on Sep 13, 2024. It is now read-only.

Visualize sample images from a plate in plate map view #8

Open
shntnu opened this issue Apr 24, 2020 · 2 comments
Open

Visualize sample images from a plate in plate map view #8

shntnu opened this issue Apr 24, 2020 · 2 comments

Comments

@shntnu
Copy link
Member

shntnu commented Apr 24, 2020

Example:

The cell counts for this plate are indeed very low by eye, just looking at the plate.
Screenshot below is one randomly selected field (3) from each well, then all the wells laid out as they would be on the plate- black is background, cells are a mix of red, green, and cyan.
You'll see that by eye >1/2 the wells are almost or completely black.

image

@shntnu
Copy link
Member Author

shntnu commented Apr 24, 2020

We may attempt to do it in cytoplot but meanwhile @bethac07 may be able to pen down ImageJ / FIJI instructions to do so

@shntnu
Copy link
Member Author

shntnu commented Apr 24, 2020

Beth Cimini documented a general protocol for making plate montages in FIJI-

  1. Have images in same general system as FIJI (ie if the images are on i_a, connect to that, if they are in s3, start a Windows VM and download the images and FIJI on that. (If doing this, make sure to grab an instance with >20GB of memory if you want to load all 6 channels))

  2. Go to FIJI -> File -> Import -> Image Sequence, navigate to the folder and select any image inside it (doesn't matter which one)

  3. In the "File name contains", put in some filtering stuff- at a minimum, you want to select one field (ie for 3 you'd put f03; if you want to get fancy and only load in channels 1, 2, and 5, you can put in (f03p01-ch[125])). The number of images remaining (ie 180 below) should be a multiple of the number of wells plated, usually 308 or 384 in a 384 well plate.
    image

  4. Press ok. Loading will take a few minutes, go check your email or grab a coffee.

  5. Go to Image-> Hyperstack -> Stack to Hyperstack, and adjust the planes and channels to be correct. It will yell at you if your math is wrong. Keep order as default, and set display to composite
    image

  6. If the first image doesn't really have cells, scroll the "Z" until you find one that does. Go to Image-> Adjust -> ColorBalance, and adjust the relative brightnessess of the channels.

  7. If you find the previous step hard, or you want to change the display color of each channel, go to Image-> Color -> Channels tool. You can then change the mode to "Color"; once you've done that, select a channel with the check boxes to change it's look-up-table (display color) under "More" or to adjust its brightness. Once all the channels look good, go back to composite mode; you can keep adjusting the brightness/contrast there to help with balance if you need.
    image

  8. When it looks good, go to Image -> Stack -> Make Montage to make the arrayed images. If your memory can handle it, keep the scale at 1.0; I like to add a 1 pixel border but that's just me. The plated rows and columns should be used to set the row and column values (usually 14 rows 22 columns or 16 rows 24 columns)
    image

  9. Once you have your montage, your color and channel menus will still work, so if it seems like some stuff is blown out you can keep adjusting until everyone looks nice.
    image

  10. Go to File -> Save to save as a TIFF, or File -> Save As to save as PNG, JPG, etc. I will often do a PNG here.

David Stirling has a semi-automated FIJI script for this (private)
https://github.com/broadinstitute/DavidStirling_Projects/blob/master/ImageJ%20Macros/Adipo_Montagemaker.ijm

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant