ľ¹Ï¸£ÀûÓ°ÊÓ

It’s now common knowledge that tobacco causes cancer, and in the last couple of decades, and this has generated efforts to try and reduce that exposure. More recently, BPA, a chemical widely used in plastics such as baby bottles and food packaging, has been associated with a variety of health hazards, which also triggered regulatory changes. Similarly, the health effects of air and water pollution has attracted a lot of attention and led to significant changes in industry standards and waste regulations.

These are several established (causal) links between health and acute exposure to certain chemicals, but the effect of our environment and everyday exposures remains unclear. How can we improve our knowledge and methodology to be able to capture molecular changes that are truly related chronic exposure?

What is the biological effect of swimming, for example?

It’s almost ironic that while you’re doing something we consider a healthy activity, you may, because of the conditions, be exposing yourself to chemicals, such as disinfectants and their by-products. We conducted a study in Spain and observed, even after a mere 40 minutes of swimming, changes at the protein level, in gene expression and metabolites (byproducts of chemical processes within our cells).

It’s almost ironic that while you’re doing something we consider a healthy activity such as swimming, you may, be exposing yourself to chemicals, such as disinfectants and their by-products.

While, it was technically impossible to determine whether the changes were due to the swimming and any swimming-related factors such as physical activity, or to exposure during the swim, we could interrogate publicly-available data and published literature. We identified, across the identified markers of the swimming experiments, some chlorinated compounds, which were more likely to be related exposures than the act of swimming itself. This is a nice example of how we can use existing knowledge to validate data that we’ve just generated and interpret results we infer.

Bridging health, exposures and methods

I’m a statistical bioinformatician, which means that I have a background in methods and a taste for biology. My group devises computational methods to better capture and understand sequences of biological alterations caused by external stressors and ultimately affecting health. We specialize in statistical modeling and are a mixture of epidemiologists, quantitative scientists with advanced modeling interests, a clinician and statisticians. We’re a go-between group, which reflects the nature of the exposome field. This buzzword is catalyzing new ways of integrating methods from different fields from mathematics to toxicology, and data from different domains, including molecular biology and social sciences. Our goal is to exploit as this wealth of information efficiently as possible and to translate it into relevance to public health and actionable results. 

The Exposome Hub

Our ongoing quest is to find a way to push our capacity up another notch to address complex questions. Can we gather more and refined information on the external stresses we are subject to? Can we explore the molecular mechanisms they trigger and explain the biological consequences, after both acute and long-term exposure? Ultimately, we aim to understand how these exposures, their biological response and their downstream consequences contribute to define individual risk profiles, which will hopefully lead to better risk stratification and health outcome prediction. In addition, the hub stresses the importance of mining existing data, in order to validate, improve or even discover previously unconsidered connections between health and their molecular origins.

Because we’re measuring exposures and biological characteristics at the individual level, we are involved in a big data enterprise that we need to sift through. The Utrecht Exposome Hub has all the components necessary to do this: data science, statistics, mathematics, exposure sciences, epidemiology, molecular biology and bioinformatics. Our ability to integrate and mine big data is critical for generating new hypotheses, improving interpretability, and therefore optimizing the impact of our research.

Imperial College London
Utrecht ľ¹Ï¸£ÀûÓ°ÊÓ