Focus on... Johanna Lepeule and the causal analysis in epigenetic epidemiology

on the March 13, 2020

Johanna Lepeule is an epidemiologist at IAB. She is working on the Developmental Origins of Health and Disease (DOHaD). She obtained a funding from the Grenoble Alpes Data Institute for her project “causal analysis in epigenetic epidemiology”.
The research project focuses on the consequences of tobacco exposure on children. It is now well known that the maternal tobacco smoking has an influence on baby and child health. It modifies the birth weight and increases the risk of premature birth, chronic diseases and mortality.

Environment (chemical such as pesticides, physical such as air pollution, tobacco smoking, way of life, etc.) may affect health early in our life, even the intrauterine life.  The biological mechanisms leading to these health effects of the environment are not well understood, but epigenetic could be one of them. Epigenetic is based on 3 mechanisms that can influence the expression of our genes without affecting their sequence. The methylation of DNA is one of them. 

Smoking or passive smoking of the mother influences the baby’s birth weight and is likely to influence the methylation of the DNA. Researchers would like to know whether the effects of smoking on birth weight are mediated by placental DNA methylation (i.e. whether maternal smoking affects DNA methylation of the placenta, which will in turn affects the birth weight of the baby). This question can be addressed using statistical mediation methods. However, there are more than 500 000 methylation marks on the genome and such a huge amount of factors is a challenge for biostatistics. Currently, no statistical method allows to reliably identify the relevant mediators among such a large number of possibilities.

On June 2017, The Grenoble Alpes Data Institute organized a data challenge on “epigenetic high-dimension mediation”. It gathered 30 researchers and students from Europe and USA and with different backgrounds. The objective of the challenge was to propose solutions and evaluate statistical methods for mediation analysis. It was the first step of this research project. It makes Johanna Lepeule (IAB) and Olivier François (TIMC- BCM) work together and obtain funding to start the research “causal analysis in epigenetic epidemiology”. Thanks to the Grenoble Alpes Data Institute, they recruited Basile Jumentier in internship. This project was successful and they obtained an ANR funding of 500 000 € that allow them to continue the project with Basile Jumentier in thesis. He develops new statistical methods to address the issue of mediation using high dimension data. He will then apply it to data from a cohort of mothers and children (EDEN, recruited from the CHU Nancy-Poitiers from 2002 to 2006). These new methods will be applied to understand whether DNA methylation in the placenta is a mechanism by which maternal smoking during pregnancy could affect the birth weight of the baby, children's respiratory health or neurodevelopment.


Article:
(1)    Caye, K., Jumentier, B., Lepeule, J., & François, O. (2019). LFMM 2: fast and accurate inference of gene-environment associations in genome-wide studies. Molecular biology and evolution, 36(4), 852-860.

Published on March 13, 2020