Common Conventions

This page is intended to serve as the place to put informaiton about depositing information in general, and extends to more than simply using the MESSES package. Some of the information will be highly specific to certain repositories or certain fields, but it’s important to get this information somewhere.

Metabolomics Standards Initiative (MSI)

The Metabolomics Standards Initiative (MSI) is an international community effort launched in 2005 by the Metabolomics Society to establish consensus-based reporting standards for metabolomics data, ensuring that experimental results are transparent, reproducible, and reusable. One of the most significant products they have produced is the minimimum reporting standards for chemical analysis. It has evolved over the years, but is mainly assigning a numerical level to an identified chemical to give it a level of confidence in its assignment. For example, a level 1 designation would be the highest confidence that the identified chemical is assigned to the correct name, and higher numbers would be less confidence. At the time of writing (April 10, 2026) there are currently 4 significant papers describing the confidence levels:

Sumner: Original guidelines from Cheimcal Analysis Working Group (CAWG), MSI. https://rdcu.be/e6DKX 2007 4 levels

Jeon: https://doi.org/10.1021/tx300457f 2013 4 levels, but is more concrete than the original paper.

Schymanski: https://doi.org/10.1021/es5002105 2014 5 levels, has level 2a and 2b

Schrimpe-Rutledge: https://doi.org/10.1007/s13361-016-1469-y 2016 5 levels

The Schymanski paper in 2014 is the most widely cited and used version of this system. The reason this system is being mentioned here is because these levels are very good information to deposit with any data it is relevant for (mass spec and NMR), but there are some issues in trying to do that deposition well. Since there are multiple versions of the system, simply depositing the numeric level associated with any chemical assignments in a dataset would be ambiguous. You also need to include which version of the system is being used. To do this we propose identifying the version by the publication year. So for example if you have something like “msi_level” in a data table you could add the column “msi_level_type” and give it the value “MSI_2014” or “MSI_2016”, whichever is the correct year being used. Since this value is unlikely to change for assignments in a given dataset, it is unlikely that multiple MSI confidence level systems are being used in one dataset unless it is combining many datasets together, a column in a table might be overkill. If there is somewhere more appropriate to indicate this within the deposition format, then it can be indicated there instead since this is more metadata than data.