Using Software To Uncover Historical Mass Spectrometry Insights

FAIR data is used more as a term in the scientific world, but this is also a problem that affects all companies. Data, if you want to use it, should be findable, accessible, interoperable, and reusable (FAIR).

Modern analytical instrumentation means that scientists acquire so much data, which is never used again because of the sheer amount. You need experts to analyze data, and it’s often hard to revisit historical data. Now, developing tools to make this data easily accessible and interoperable—especially with different vendor systems in each lab that we see in mass spectrometry—is a key part in moving science forward.

Our project aims to make this approach ready for enterprises now. The key here is intellectual property, but also the fact that most of the public data comes from natural products analysis or metabolomics—very specific scientific topics which may not be relevant for a chemical company. Also, these companies don’t really want to publicly share what they are working on. So, creating a platform following the FAIR principles to make data easily accessible, and in the end, very reusable in an enterprise environment, is the key that will help to push innovation in these companies.

The software is going well in development at the moment, and we plan to release it this year under a different name—so FAIR-MS is a working name for the project funded by the €150,000 grant we have just received from the city of Bremen.

In an ideal world, if a researcher analyzes new data where maybe five years ago another scientist analyzed something similar, our tool should be able to automatically find that and highlight that. If left up to a human, nobody would go back into five-year-old data and manually crunch it. This tool makes data more easily findable, and therefore reusable, and with this additional knowledge, you are really about to leverage what you have invested in the past.

We believe that there is so much data being created, but that lots of it is lost. Once a project is done, its data is archived, but will anybody ever go back to it and reuse it? This data is a goldmine that people are sitting on.

From a sustainability perspective, measuring one sample with mass spectrometry does require a lot of energy and chemicals—a lot of high-purity chemicals. Our ambition is to help get the most out of each sample and every measurement.

In topics such as lipidomics—especially with oxidative stress and oxidative lipids—the analytical challenge there is extremely difficult, even with super-high resolution or ultra-high resolution mass spectrometry. The nuances there are very, very tiny. The more accurate you go, the better your results will be. There is also a high chance of, for example, doing a false-positive lipid identification using multi-modal techniques like ion mobility, liquid chromatography with mass spectrometry, and also doing it spatially. This helps the biological interpretation, but it also helps the overall identification.

These different techniques usually also require different standards being used for calibration. With our partners at Polymer Factory in Sweden, they are now developing new chemical standards, and we want to incorporate this as one of the standard functionalities in our software to make it very easy for the final user to use this complete solution. Our partners in Mannheim, the group of Prof. Dr. Carsten Hopf at CeMOS (Center for Mass Spectrometry and Optical Spectroscopy), is now developing the analytics and will really put to the test [the software] that we are now developing. With their feedback and input, we hope to tackle this huge scientific and analytical challenge of oxidized lipids.

This project—LIMMIC—has just started. But our partners at Polymer Factory already have some new prototypes, and we are now implementing the changes to our software to make it available in the coming months to our partners. 

We have brought our latest release of mzmine here with us at ASMS—we have released version 4.10—and in this, we have made lots of great steps in usability and ease-of-use. We are bringing in compound and lipid dashboards to make analysis even easier.

One very notable change is that, in our field, people like to analyze features, so we are now taking a step towards more compounds. We are now, for the user, grouping features into compounds to help them identify different adducts, isotopes, and in-source fragments to further simplify the dataset and give them the ability to focus more on the science itself.

We are very excited to see how the community will react to it—we currently have more than 20,000 users, especially in academia, and we have a very fast feedback loop. Maybe within the next week or so, we will see the first comments and feedback.