Skip to content


Repository files navigation



This Javascript can only work when called by imageJ/Fiji. The GUI version must be run within the Fiji script editor. The CLI version must be run with xvfb-run in a terminal session.

Both scripts create a montage cell-grid image where cells identified in microscopy images (saved as ROIs) are extracted and arrayed on a grid for each well of a 384-wellplate.

Format of the screen definition file

To apply our script to your images, a tabulated file (TSV or CSV format) with the following fields (columns) is required. The first row should start with a hashtag sign ("#") followed by the name of the required fields :


  1. plate gives the number of the 384 well plate from which images were acquired (for a proteome-wide screen, this typically ranges from 1 to 16).

  2. well id must be 3 characters long starting from A01 to P24 (it should be 384 wells exactly).

  3. orf must contain an uppercase identifier for the gene locus tagged.

  4. c0 must contain full path to the images containing segmentation information. ROIs must be saved as overlays and the images should be in OME-TIFF file format.1

  5. c1 must contain full path to the images corresponding to the first fluorescent channel.1

  6. c2 must contain full path to the images corresponding to the second fluorescent channel (if available, set to NA otherwise).1

  7. c3 must contain full path to the images corresponding to the third fluorescent channel (if available, set to NA otherwise).1


#plate well orf c0 c1 c2 c3
plate1 A01 YAL002W screen-brightfield-00001.tif screen-GFP-00001.tif screen-RFP-00001.tif screen-BFP-00001.tif
plate1 A01 YAL002W screen-brightfield-00002.tif screen-GFP-00002.tif screen-RFP-00002.tif screen-BFP-00002.tif
plate1 P24 NA NA NA NA NA
1 The fields corresponding to channels (c0,c1,c2,c3) must contain full paths to the images and should correspond to the plate, well and orf written. If several images were acquired per well, you can add as many rows as needed as long as the "plate", "well" and "orf" fields are duplicated (cf example above).If you have missing images, the fields may be left blank or indicated by "NA" (i.e. not available).


initialize ImageJ
STATUS  = initialize global variables
MONTAGE = set Montage Parameters
FIELDS = read and split first row of **input.tsv**
if FIELDS corresponds to : plate well orf c0 c1 c2 c3
    INPUT = load **input.tsv**
    exit("Input file is invalid.")

For each CHANNEL ( c0 c1 c2 c3 )
    CHIMG = open image from CHANNEL
    For each rowInput in INPUT:
        1. Check if well has been visited
           if FALSE :
              update STATUS (picnum reset to 1, get well position on plate)
              GRIDIMG = create grid image for each well

        2. Get detected cells in segmented image (channel c0) in Array of Cell Objects (CELLS)
           if CELLS length is 0 :
              skip to next rowInput
              if STATUS->Copy == TRUE
                  3.a) copy ROI into CELLS from CHIMG to corresponding GRIDIMG
                  if STATUS->CopiedCells == MONTAGE->MaxCellPerWell
                     save GRIDIMG as TIFF with labels
                     set STATUS->Copy to FALSE
                  3.b) skip to next rowInput    
    end of INPUT loop
end of CHANNEL loop

If you wish to upload your own microscopy screens to, please follow these guided instructions. Implortantly, the image processing script will require ALREADY identified cells or cell centers in your images (saved as ROIs). 

Send an email at with the following information :

  • Corresponding email address to be mentioned on the web-site,
  • Authors names,
  • Screen name (less than 15 characters),
  • Publication reference(s)
  • Strain genotype and condition in which it was imaged,
  • Microscopy screen definition file (detailed above)


Script for "cell-grid montage" of yeast microscopy images






No releases published


No packages published