Examples of public MSI datasets for SpacePath rapid testing

This record contains example datasets, of small size, for a fast test of SpacePath, a computational framework for pathway and molecular family analysis for MSI-based spatial metabolomics and lipidomics.

Instructions:

The working directory can be your 'home' or the location of your preference in your system:

• Download, from this Zenodo record, the examples-sm-data.zip file and unzip it. 
• Place the unzipped `examples-sm-data/` inside your working directory:
  • at the same level as your SpacePath local clone, if using this approach, or
  • at the same level of the folder containing the parameters files, if installed SpacePath via 'pip'.
• Also in the working directory, create a `figures/` folder, where the output will be stored: cd $HOME; mkdir figures

the final expected structure is:

👉 Please visit the Wiki page of SpacePath which provides documentation for running the pipeline on this example data.

Explanation of the provided example datasets

Two datasets of spatial lipidomics/metabolomics MSI data are available in this compressed folder, within two subfolders:

• example-cerebellum/: the data from Bonjd N, Griffin J, Hall Z et al., 2017 
• example-corn-root/ : the data from Evan Larson, 2021 (publicly accessible in Metaspace, no paper associated)

Each subfolder contains two annotations files (HMDB and LipidMaps) in the form of .csv files,  and the unified intensities into a single 'AnnData' object in .h5ad file format. 

Also, tables for metabolites mappings are provided inside the same compressed folder, in the subfolder 'metabolites-mappings' which are used by SpacePath (for identifiers conversions and compound-to-pathway mappings). These tables were generated using the code available at https://zenodo.org/records/23079160.

Acknowledgements:

Acknowledgements to the authors of the two datasets in this folder:

• example-corn-root: Evan Larson, 2021, at https://metaspace2020.org/dataset/2021-02-10_20h21m12s
• example-cerebellum: Bonjd N, Griffin J, Hall Z et al., 2017, at: https://www.ebi.ac.uk/metabolights/editor/MTBLS487/assays

We also thank Metaspace https://metaspace2020.org where the intensities and annotations can be retrieved with the registires 2021-02-10_20h21m12s and 2025-04-10_14h53m38s respectively (https://metaspace2020.org/dataset/[registry]). 

The datasets were downloaded using the `metaspace-converter` Python package, used within a custom script to unify the intensities into a single anndata object for each dataset.

----

Note: the classification of spots, using SpacePath auxiliary scripts,  has already been applied to the objects (.uns and .obs attributes), and saved to the .h5ad file.