MS_targeted is an open-source command-line pipeline for statistical analysis of mass spectrometry metabolomics data. The pipeline is implemented in C# and R, and runs in all platforms. In Windows you can run it on cmd.exe, while in OSX and Linux on the terminal with the cross-platform open-source .NET framework mono. MS_targeted can handle multiple tissues and charges simultaneously.
For any questions or issues please use the Issues in github or contact Klev Diamanti.
- You will find the pre-compiled executables under MS_targeted/exec/.
- It is required that you have installed the latest version of mono and R.
- The R packages lmPerm, coin, gridExtra, ggplot2, Hmisc, RcmdrMisc, RVAideMemoire and lm.beta are required to be pre-installed.
- You will need to decompress the database file prior to running the pipeline.
- Detailed explanation of the input files and options is provided in MS_targeted_input.md.
- Detailed explanation of the output files is provided in MS_targeted_output.md.
[mono] MS_targeted.exe input_ms_dir clinical_data_file combined_db_file conf_file output_dir
Prior to running MS_targeted you will need to structure the metadata (covariates) and mass spectrometry input data according to the sample files from metabolomics workbench study ST000383 (Fiehn et al., 2010).
Directory where the input mass spectrometry data are stored in plain text comma- or tab-separated files. All the input files should have the same prefix prior to the first underscore (_) and the extension. The various database id's for each metabolite should be set in the input files. We recommend you have one file for each tissue. Please check MS_targeted/sample_data/input_ms/ for an example. Note that one custom id for every metabolite is required, that is defined in the row m_id. These id's should contain at least one letter and numbers.
File where the metadata or covariates for the samples are stored. This is a plain text comma- or tab-delimited file. The first row should contain unique names for each covariate. The next row should contain the type of the covariate (exclusively categorical or numeric). Please check MS_targeted/sample_data/metadata/ for an example.
This is a meta-data file that contains various database id's for thousands of metabolites. The file should be the output from the repository metabolomicsDB. For a start you might use the tab-separated file under MS_targeted/sample_data/db/. Please decompress the file before using it.
This file configures various parameters for MS_targeted. The sample file under MS_targeted/sample_data/conf/ contains comments and details.
An existing or not existing output directory where all the output files will be written. Please note that if the directory exists it will be overwritten.
Klev Diamanti, Marco Cavalli, Gang Pan, Maria João Pereira, Chanchal Kumar, Stanko Skrtic, Manfred Grabherr, Ulf Risérus, Jan W Eriksson, Jan Komorowski and Claes Wadelius (2019). "Intra- and inter-individual metabolic profiling highlights carnitine and lysophosphatidylcholine pathways as key molecular defects in type 2 diabetes". Scientific Reports 9(1):9653.