The napari-bigfish plugin

Start napari and run Detect FISH spots (napari-bigfish) from the menu Plugins. Open the image containing the FISH-spots. If you also want to count spots per cell and per category nucleus/cytoplasm, open an image with the cell labels and one with the nuclei mask or labels.

The plugin comes with an example image that you can open via File>Open Sample>BigFISH smFISH Analysis (napari-bigfish). This is a synthetic image that contains the FISH-Spots in the background and in 4 “cells” each one with one “nucleus”.

When using your own images, make sure that the scale in the image (pixel size in x, y and z) is correctly set. The plugin requires the scale to be in nanometers (nm). If you use the naparj-j plugin to transfer images from FIJI/ImageJ to napari, you can set the scale in the properties dialog available via Image>Properties....

In napari you can check and modify the scale attribute of a layer from the integrated IPython console:

In [1]: viewer.layers.selection.active.scale
Out[1]: array([1000.    ,  108.3424,  108.3424])

Background Subtraction

The operation removes the background by subtracting a blurred version of the image from the original image. A new image layer with the result image will be created.

sigma xy

The standard deviation of the Gaussian kernel in the xy-plane.

sigma z

The standard deviation of the Gaussian kernel in the z-dimension.

See also:

bigfish.stack.remove_background_gaussian

Spot Detection

The operation uses a LoG-filter and a local maximum detection to detect spots. A new points-layer containing the detected spots will be added to the viewer.

threshold

The local threshold used for the spot detection. If find threshold is selected, this value will be ignored and the auto-detected threshold value will be used instead. The field will be updated with the detected threshold value.

spot radius xy

The radius of a spot in the xy-plane in nanometer.

spot radius z

The radius of a spot in the z-dimension in nanometer.

remove duplicates

Remove potential duplicate coordinates for the same spots. The option slows the execution.

find threshold

If selected the threshold is automatically detected, otherwise the value from the threshold-field is used.

See also:

bigfish.detection.detect_spots

Decompose Dense Regions

The operation breaks up clustered regions by building a reference spot and fitting it as many times as possible into the region. A new points-layer with the resulting spots will be added to the viewer.

spots

The points layer of the spots from a previous detection to which the decomposition shall be applied.

spot radius xy

The radius of a spot in the xy-plane in nanometer.

spot radius z

The radius of a spot in the z-dimension in nanometer.

alpha

Intensity percentile used to compute the reference spot, between 0 and 1. The higher, the brighter are the spots simulated in the dense regions. Consequently, a high intensity score reduces the number of spots added. Default is 0.5, meaning the reference spot considered is the median spot.

beta

Multiplicative factor for the intensity threshold of a dense region. Default is 1.

gamma

Multiplicative factor used to set the gaussian kernel size, for a gaussian background removal. A large gamma increases the scale of the gaussian filter and smooth the estimated background. To decompose very large bright areas, a larger gamma should be set.

See also:

bigfish.detection.decompose_dense

Spot Counting

Counts the spots in the image. If a label image for cells is given the spots are counted per cell. If in addition a label image or a binary mask of the nuclei is given the spots within the cytoplasm and the nuclei are counted separately per cell. A table containing the results will be opened in the viewer. The results from the table can be exported to a spreadsheet program using copy and paste.

spots

The previously detected spots in the form of a points-layer.

cytoplasm labels

A labels layer with labels of the cells. The background should have the label 0.

nuclei labels of mask

A labels layer in which the background has the label 0 and voxels belonging to a nucleus have a value bigger than 0.

To select the whole table click into the upper-left corner of the table. To copy the selected data use either ctrl+c or the contex-menu of the table.