How can mathematics help us to understand and combat obesity? In the MATOMIC (Mathematical Modelling for Microbial Community Induced Metabolic Diseases) project, computer scientist Professor Daniel Merkle and his interdisciplinary team are deciphering the interactions in the intestinal microbiome and making predictions about intestinal processes. The project led by Merkle is funded by the Danish Novo Nordisk Foundation.
How does the microbiome influence our health and how can it be used to treat obesity?
Daniel Merkle: The human microbiome—that is, the entirety of microorganisms in the body—has been shown to have an enormous influence on our health. Our intestines, for example, are home to thousands of bacteria that are essential if our body is to function properly. The microbiome is altered in obesity. One therapeutic approach is fecal microbiome transplantation (FMT): this involves transferring the microbiome of a slim person to an obese person. All previous FMT attempts have failed due to incompatibilities between the microbiomes of donors and recipients that impair stability. At MATOMIC, we are working to improve our understanding of these processes by applying mathematical modelling.
© Bielefeld University
So, you translate processes in the gut into mathematical language?
Daniel Merkle: Exactly. At MATOMIC, we try to translate chemical reactions into mathematical descriptions. We then analyse which processes we can reproduce with the modelled reactions. Our model of these reactions then helps us to predict what actually happens in the body at a chemical level. One example of such a chemical reaction: when sugar is ingested, specific bacteria process the carbon it contains, pass it on to other bacteria, and thus influence major processes in the body.
How can your research findings be used in medicine?
Daniel Merkle: Autoimmune diseases are strongly influenced by the microbiome. In our modelling, it is not important whether we are addressing asthma, obesity, or other diseases. The focus is on personalised medicine, which is not yet established. Every person has a unique microbiome, and this interaction needs to be better understood.
And how exactly do you gain this understanding?
Daniel Merkle: We apply methods from computer science to better understand the processes in the microbiome. You can think of it like a map of chemical processes. Graph grammars are a key tool here. They enable us to visualize chemical reactions as a network of compounds. This makes it possible to understand how substances are metabolized in the body. In addition, we use special analyses to find out which processes have which effects—in other words, which chemical changes lead to specific reactions in the body. Models that record parallel chemical processes are another tool. For example, one member of our team is using network analyses to investigate how different bacteria in the intestines interact.
© Bielefeld University
You are currently investigating eight bacterial species in the bioreactor, but there are thousands of them in the real world. How do you deal with this discrepancy, and is it in any way possible to make a prediction?
Daniel Merkle: That’s one of the key questions we were asked when we were defending our project. That is simply research: we have to try it out. If we can’t understand the processes with eight species, then we certainly won’t be able to decipher them with 8,000. The idea, of course, is to start with small models and then transfer our findings to larger systems.
How do you check the accuracy and reliability of your results?
Daniel Merkle: We combine our modelling techniques with experimental cultivations of microbiomes of varying complexity. In Leipzig, the team led by Martin von Bergen from the Helmholtz Institute there is investigating how the eight bacterial species interact in the intestines. Parallel to this, Peter Stadler from Leipzig University is analysing the microbial species, while Christoph Flamm from the University of Vienna is using dynamic modelling techniques to analyse temporal interactions. In this way, we are continuously refining our models and improving their predictive power.
So, you find yourself in a permanent feedback loop?
Daniel Merkle: Exactly, we meet up in person three times a year and have virtual meetings every month. There is a lot of discussion. But that’s important for joint research. You can’t be too shy to ask questions that might be easy for scientists from other disciplines to answer.
Interdisciplinarity is therefore essential for your research.
Daniel Merkle: The move from the University of Southern Denmark to Bielefeld a year ago came at the perfect time for me. In Denmark, I was missing some of the interdisciplinary collaboration that is practised intensively here at the university. Since I started in December 2023, I have made valuable contacts—for example, during retreats at CeBiTec and discussions with physicists, mathematicians, chemists, and biologists. This scientific exchange is very enriching for me: I see my future here in Bielefeld.
TACsy
The Training Alliance for Computational Systems Chemistry (TACsy) is another major EU-funded project headed by Professor Dr Daniel Merkle. The next generation of scientists is being trained here, with a strong emphasis on interdisciplinary collaboration between computer science, chemistry, and machine learning. The project focuses on application areas such as CO2 recycling, enzymatic synthesis design, and lipidomics. Doctoral students are collaborating with several research groups and industrial partners to develop methods and transfer them to industry.
