Metabomatching is a procedure to identify compounds from spectral features that jointly associate with genotypes.
- The simplest method to run metabomatching is to use its implementation in the PhenoMeNal portal public.phenomenal-h2020.eu, located in the NMR section. Because the PhenoMeNal implementation is designed for ease of use, some advanced features may not be accessible. A set of test data is provided.
- Alternatively, the metabomatching docker image can be pulled and run locally. With a working docker installation (get docker), run
docker pull docker-registry.phenomenal-h2020.eu/phnmnl/metabomatching
to download the metabomatching container. The download may take some time, as the container includes a full Linux OS, octave to run metabomatching, and inkscape to convert metabomatching figures from SVG to PDF. To run metabomatching on the included sample pseudospectrum, run
docker run -i -v <absolute/path/to/my/directory>:/mm-ps metabomatching/metabomatching
which will run metabomatching, and write results to
<absolute/path/to/my/directory>. Note that on Windows, the absolute path must be formatted without using a colon, so to use the directory
C:\MyDir, the path must be formatted as
- To run metabomatching on your own pseudospectra, please refer to the documentation (PDF).
- To run metabomatching directly, fully featured, and fully up to date (on a system running either Matlab or octave), metabomatching can also be obtained from GitHub
The procedure was in applied in
- the metabolome-wide genome-wide association study in the CoLaus cohort PLOS Genet, 10(2): e1004132
- the metabolome-wide genome-wide association study in the SHIP cohort PLOS Genet 11(9): e1005487