This repository is for developing and evaluating various implementations of single-node spatial metabolomics pipelines. As such, it's not at all production ready and can easily eat your cat^W^Wall available RAM.

Having said that, here's how to install and run the pipelines:

• Install the necessary Python modules: pip install -r requirements.txt
• (Optional, for faster measure_of_chaos) install OpenCV 3.0
• Copy the example.json template and edit it as necessary:
  • method can be either reference or new
  • file_inputs→data_file must point to the input file
    • 'reference' pipeline accepts only the *.hdf5 format
    • 'new' pipeline accepts both *.imzML and *.hdf5
• Run the pipeline: python2 pipelines.py your_config.json
• The following results are stored in file_inputs→results_folder:
  • a file with scores for all formulae;
  • a file with scores only for the formulae that pass the thresholds
  • images for each (formula, adduct) pair that passes the filter

Funding

This project is funded from the European Horizon2020 project METASPACE (no. 634402), NIH NIDDK project KPMP and internal funds of the European Molecular Biology Laboratory.

License

Unless specified otherwise in file headers, all files are licensed under the Apache 2.0 license.