Mots-Clés
Heat stress
multi-omics integration
DNA methylation
transcriptomics
chicken
biomarkers
Description
Context
Climate change is one of the major challenges that animal production is facing worldwide. Hot climate extremes are increasing not only in tropical regions but also in temperate areas.
Chickens are highly vulnerable to heat stress, due to their poor thermoregulatory capacity due to feather insulation and the absence of sweat glands. Heat stress acts through complex physiological changes and epigenetic mechanisms as the DNA methylation, appear to mediate physiological responses and adaptation.
High-throughput technologies now enable comprehensive analysis of multiple layers of molecular information.
Aim of the thesis
This thesis will make use of a unique dataset combining transcriptomic and DNA methylation data obtained from five chicken genotypes exhibiting contrasting levels of heat resistance and feed efficiency, either exposed or not to heat stress.
The aim is to apply and evaluate statistical methods for the integrative analysis of high-dimensional multi-omics data, and to better understand the molecular basis of heat resistance. This includes the use of existing integrative statistical approaches (e.g., MOFA, DIABLO, Bayesian models, network-based methods), the assessment of their relevance and limitations in this context, and the testing of new methodologies developed within the research group.
The results will then be basis for identifying stress-related molecular signatures that are either shared across genotypes or specific to ones, and to link these markers to the extensive set of available phenotypes (production, feed efficiency, quality, physiology).
The ultimate goal is to identify predictive biomarkers of thermotolerance and to improve our understanding of the biological mechanisms underlying resistance or sensitivity to heat stress in laying hens.
The questions we will try to answer are:
-
What is the role of methylation in liver and blood in the response to heat stress?
-
Are methylation changes comparable among tissues?
-
Do environmental factors induce methylation changes that correlate with expression differences?
-
Are there genotype-specific methylation patterns that mediate expression responses (GxE at the epigenetic level)?
-
To what extent are methylation and transcriptome related to the variability of phenotypic traits?
Overview of the statistical methods considered
Statistical analysis methods could include, for instance:
-
differential analysis of omics data to identify Differentially Expressed Genes (DEG) (e.g. with R packages DESeq2) and Differentially Methylated Cytosines and Regions (DMC and DMR) (e.g. with DSS) in order to identify an initial list of candidate markers and for subsequent gene set enrichment analyses;
-
differential methylation analysis at the gene level (detecting changes in the full shape of their methylation profile) and at the region level (promoter, gene body, …), will also be explored in order to better understand the link between methylation and gene expression;
-
multi-omics integrative analyses to understand the between-omics relations, by using unsupervised (e.g. multiple factor analysis such as MOFA) and supervised approaches (e.g. multiblocks sPLS-DA such as DIABLO), and subsequently link these patterns with the available phenotypic traits;
-
Clustering, functional and gene set enrichment analysis (WGCNA and GSEA, for example with clusterProfiler) to identify biological paths implicated in the heat stress resistance (or sensitivity);
-
identification of heat stress (sensitivity) markers using machine learning methods such as regularized (logistic) regression, or random forests after an initial pre-selection of the potential markers;
-
methods recently developed within GiBBS team (GABI, INRAE) for multi-omics analyses, such as:
-
BayesOmics¹ (taking into account correlation within omics),
-
Idiffomix² (methylation and gene expression altogether - from paired samples),
-
Heterocop³, to infer multi-omics correlation networks (after a first pre-selection of candidate methylation sites and genes).
¹ https://github.com/terenceviellard/BayesOmics
² https://cran.r-project.org/web/packages/idiffomix/vignettes/vignettes.html
³ https://github.com/cran/heterocop/tree/master
Work environment
The PhD student will be part of the GiBBS team and will be supervised by Tatiana Zerjal, Marie Courbariaux, and Andrea Rau. The thesis will also benefit of the expertise of Florence Jaffrezic in statistical approaches for data integration.
The project is embedded in the international research consortium built in the framework of the “GEroNIMO” with established collaborations with the University of Wageningen and Uppsala University.
Desired profile and skills
We are seeking a highly motivated candidate with a good background in (bio)statistics. The applicant should also have a solid understanding of the biology underlying adaptation to environmental stress.
Good programming skills in R and experience in data analysis are required. A background in bioinformatics would be a plus.
The candidate should demonstrate analytical and critical thinking skills, the ability to work both independently and collaboratively, and good communication skills in English.
A strong interest in interdisciplinary research at the interface between biology and data science is essential.
Bibliography from the team (a sample)
Karami, K., Sabban, J., Cerutti, C., Devailly, G., Foissac, S., Gourichon, D., Hubert, A., Hubert, J.-N., Leroux, S., Zerjal, T., Lagarrigue, S., Pitel, F. (2025). Molecular responses of chicken embryos to maternal heat stress through DNA methylation and gene expression: a pilot study. Environmental Epigenetics 11, dvaf009. https://doi.org/10.1093/eep/dvaf009
Majumdar, K., Jaffrézic, F., Rau, A., Gormley, I. C., & Murphy, T. B. (2024). Integrated differential analysis of multi-omics data using a joint mixture model: idiffomix. arXiv preprint arXiv:2412.17511.
Coustham, V., Andrieux, C., Cerutti, C., Collin, A., David, I., Demars, J., Devailly, G., Morisson, M., Houssier, M., Lagarrigue, S., Métayer-Coustard, S., Mignon-Grasteau, S., Panserat, S., Petit, A., Vitorino Carvalho, A., Zerjal, T., Pitel, F. (2023). Epigénétique, gènes et environnement : quelle importance pour les pratiques d’élevage et les méthodes de sélection des volailles ? INRAE Prod. Anim. 36, 7384. https://doi.org/10.20870/productions-animales.2023.36.4.7384
Jehl F., Désert C., Klopp C. C., Brenet M., Rau A., Leroux S., Boutin M., Lagoutte L., Muret K., Blum Y., Esquerre D., Gourichon D., Burlot T., Collin A., Pitel F., Benani A., Zerjal T., Lagarrigue S. (2019). Chicken adaptive response to low energy diet: main role of the hypothalamic lipid metabolism revealed by a phenotypic and multi-tissue transcriptomic approach. BMC Genomics, 20(1). DOI: https://doi.org/10.1186/s12864-019-6384-8.