|Entreprise/Organisme :||INRAE, unité NutriNeurO|
|Niveau d'études :||Master|
|Sujet :||Impact of nutritional supplement interventions on cognitive health networks|
|Date de début :||Spring 2022|
|Durée du contrat :||6 months|
|Rémunération :||Approx 580€ / month|
|Secteur d'activité :||Biostatistics, nutrition|
|Description :||The widespread aging of the population brings new problems linked to cognitive decline, which is a major determinant in dependency and quality of life in the elderly. Epidemiological studies have suggested a protective role of nutrition against cognitive decline. This beneficial effect has been attributed in part to some nutrients including long-chain omega-3 polyunsaturated fatty acids, particularly docosahexaenoic acid (DHA) and some vitamins (A, B6, B12, C, D and E), which exert anti-inflammatory and/or antioxidant effects. In this context, the objective of the NUTRIMEMO clinical study was to examine the nutritional supplementation effects of DHA and vitamin A on cognitive functions in healthy elderly people, fueling a shared interest at INRAE in understanding how nutrition impacts health status.
In the NUTRIMEMO study, a variety of parameters (cognitive and blood parameter measures, nutrient dosages) were recorded before and after nutritional intervention for approximately 300 healthy individuals 60-70 years of age, evenly split between the nutrition intervention and placebo groups. Measured variables are made up of a variety of types, including quantitative measures, counts, and ordinal scores. A central question of interest is the identification of variables, as well as interdependencies among diverse variables, for which a significant effect of the intervention can be observed as compared to the placebo.
(1) After performing exploratory and graphical analyses of the data, the intern will validate the initial hypothesis with in-depth analyses using classical statistical techniques such as (generalized) linear (mixed) models, as well as machine learning predictive models (e.g., CART, random forest);
(2) Initial analyses focused on univariate models alone and ignored significant relationships among the various mixed-type measures collected. The intern will next perform a review of available network inference approaches (e.g., Gaussian graphical models, correlation networks) that can be used in conjunction with data transformations to identify interdependencies among mixed-type measures.
(3) Finally, using the selected network inference approach, the intern will identify intervention-specific networks among cognitive scores, biological parameters, and nutrient dosage in the intervention and placebo groups.
● Solid knowledge of the R programming language for data manipulation, analysis, and visualization;
● Knowledge of and experience with linear mixed models and standard machine learning approaches. Experience with network inference approaches (e.g., Gaussian graphical models, correlation networks) would be appreciated but is not required;
● Comfortable reading/writing English (ability to read and understand scientific articles and write reports of results);
● Candidates should have motivation and interest for applications in nutritional neuroscience in general, but experience in this area is not required.
The Master’s internship will take place in a research environment that brings together biologists and biostatisticians. The research internship will take place at the Université de Bordeaux, although an alternative location at the Inrae Hauts-de-France Research Center in Estrées-Mons (80) would be a possibility for interested candidates. The work will be jointly supervised by Dr. Jean-Christophe Delpech, Dr. Charlotte Madore-Delpech, and Dr. Andrea Rau.
firstname.lastname@example.org; email@example.com; firstname.lastname@example.org|
|En savoir plus :||https://www6.bordeaux-aquitaine.inrae.fr/nutrineuro|