Skip to content

Look at the result

Your result is a folder ending in .zarr. This page shows what is inside it and how to open it.

What is inside

Part What it holds
The spectra One row per measured pixel and one column per m/z bin
Average spectrum The mean spectrum of all measured pixels
TIC image The total ion current of every pixel, as an image
Pixel outlines The position and size of every measured pixel
Optical image The photo of the slide, when the acquisition has one (Bruker)
Description The instrument, the settings and the pixel size: see Describe and share your data

There are no ion images in the result: you make them from the spectra, as shown below. Most pixels have no signal in most m/z bins, so Thyra stores only the values that are there, compressed. That keeps the result small.

Advanced: the names of the parts

Every part is named after the dataset, msi_dataset unless you choose another name with --dataset-id, and after the slice, _z0 for the first one:

  • msi_dataset_z0: the table of spectra
  • msi_dataset_z0_tic: the TIC image
  • msi_dataset_z0_pixels: the pixel outlines
  • msi_dataset_optical_...: the optical images, named after their image files

Some data adds more. Bruker timsTOF data with ion mobility carries a mass-mobility heatmap in the table, and PASEF MS/MS data adds a table ending in _msms. Ion mobility and MS/MS shows how to use them. Output format lists every part, including those for 3D data.

Keep the folder whole

The .zarr folder is one dataset. Copy it, move it or share it like any other folder, but do not take files out of it or rename files inside it. It opens on Windows, macOS and Linux alike.

Open it without writing code: napari

napari is a free image viewer. With the napari-spatialdata plugin, it opens Thyra results.

  1. Install napari with the plugin. With uv (see Install):

    uv tool install --python 3.13 napari --with "napari-spatialdata[all]"

  2. Start napari:

    napari

  3. Open your .zarr folder with File > Open Folder.... If napari asks which plugin should open it, choose napari-spatialdata.

  4. A SpatialData panel appears on the left. Click global in its list of coordinate systems.
  5. The panel now lists the parts of the result. Double-click msi_dataset_z0_tic to show the TIC image.

napari shows the TIC image, the optical image and the pixel outlines well. For an image of one m/z value, use Python, as below.

Open it in Python

A few lines of Python give you the TIC image, or an image of any m/z value. The Tutorial runs them step by step on example data.

You need Thyra in the Python you work in, and matplotlib to draw the images (see Already use Python?):

pip install thyra matplotlib

Then:

import matplotlib.pyplot as plt
import numpy as np
import spatialdata as sd

sdata = sd.read_zarr("brain_section_1.zarr")

# The TIC image
tic = np.asarray(sdata.images["msi_dataset_z0_tic"])[0]
plt.imshow(tic)
plt.show()

# An image of m/z 760.6: add up the intensities within 0.25 of it
table = sdata.tables["msi_dataset_z0"]
mz = table.var["mz"].values
lo, hi = np.searchsorted(mz, [760.35, 760.85])
values = np.asarray(table.X[:, lo:hi].sum(axis=1)).ravel()
x = table.obs["x"].values.astype(int)
y = table.obs["y"].values.astype(int)
ion_image = np.zeros((y.max() + 1, x.max() + 1))
ion_image[y, x] = values
plt.imshow(ion_image)
plt.show()
Advanced: why a window, and not one bin

On the shared m/z axis, most bins are empty. The single bin nearest your m/z value is usually one of them, so an image of it would be blank. Add up the bins in a small window around the value instead, as above.

To view the result in napari from Python, install napari-spatialdata[all] in the same environment and run:

from napari_spatialdata import Interactive

Interactive(sdata)

Go deeper