Biology subjects

Bouchard, L.

Publications and source records attributed to Bouchard, L..

3 recordsLinked to original sources

bioRxiv2017

Placenta DNA Methylation Adaptation to Maternal Glucose Tolerance in Pregnancy

Maternal hyperglycemia during pregnancy is associated with fetal growth and adverse perinatal and developmental outcomes. Placental epigenetic maladaptation may underlie these associations. We performed an epigenome-wide association study of term placentas and prenatal maternal glucose response 2-hour post oral glucose challenge at 24-30 weeks of gestation among 448 mother-infant pairs. Maternal glucose levels post-load were strongly associated with lower DNA methylation of 4 CpGs (FDR q<0.05) within the Phosphodiesterase 4B gene (PDE4B). Additionally, three other CpGs were differentially methylated relative to maternal glucose response within the TNFRSF1B; LDLR; and BLM genes (FDR q<0.05). Methylation levels correlated with expression in placental tissue for all 4 CpG sites in PDE4B (rs: 0.26-0.35, P<0.01), LDLR (rs: 0.22, P=0.03) and at TNFRSF1B (rs: -0.25, P=0.01). Our study provides evidence that maternal glucose response during pregnancy is associated with DNA methylation of genes within the placenta that are partially under epigenetic control.

genomics↗

bioRxiv2017

Maternal and fetal genetic contribution to gestational weight gain

BackgroundClinical recommendations to limit gestational weight gain (GWG) imply high GWG is causally related to adverse outcomes in mother or offspring, but GWG is the sum of several inter-related complex phenotypes (maternal fat deposition and vascular expansion, placenta, amniotic fluid and fetal growth). Understanding the genetic contribution to GWG could help clarify the potential effect of its different components on maternal and offspring health. Here we explore the genetic contribution to total, early and late GWG.\n\nParticipants and MethodsA genome-wide association study was used to identify maternal and fetal variants contributing to GWG in up to 10,543 mothers and up to 16,317 offspring of European origin, with replication in 10,660 mothers and 7,561 offspring. Additional analyses determined the proportion of variability in GWG from maternal and fetal common genetic variants and the overlap of established genome-wide significant variants for phenotypes relevant to GWG (e.g. maternal BMI and glucose, birthweight).\n\nResultsWe found that approximately 20% of the variability in GWG was tagged by common maternal genetic variants, and that the fetal genome made a surprisingly minor contribution to explaining variation in GWG. We were unable to identify any genetic variants that reached genome-wide levels of significance (P<5x10-8) and replicated. Some established maternal variants associated with increased BMI, fasting glucose and type 2 diabetes were associated with lower early, and higher later GWG. Maternal variants related to higher systolic blood pressure were related to lower late GWG. Established maternal and fetal birthweight variants were largely unrelated to GWG.\n\nConclusionWe found a modest contribution of maternal common variants to GWG and some overlap of maternal BMI, glucose and type 2 diabetes variants with GWG. These findings suggest that associations between GWG and later offspring/maternal outcomes may be due to the relationship of maternal BMI and diabetes with GWG.

genetics↗

bioRxiv2017

An analytic approach for interpretable predictive models in high dimensional data, in the presence of interactions with exposures

Predicting a phenotype and understanding which variables improve that prediction are two very challenging and overlapping problems in analysis of high-dimensional data such as those arising from genomic and brain imaging studies. It is often believed that the number of truly important predictors is small relative to the total number of variables, making computational approaches to variable selection and dimension reduction extremely important. To reduce dimensionality, commonly-used two-step methods first cluster the data in some way, and build models using cluster summaries to predict the phenotype.\n\nIt is known that important exposure variables can alter correlation patterns between clusters of high-dimensional variables, i.e., alter network properties of the variables. However, it is not well understood whether such altered clustering is informative in prediction. Here, assuming there is a binary exposure with such network-altering effects, we explore whether use of exposure-dependent clustering relationships in dimension reduction can improve predictive modelling in a two-step framework. Hence, we propose a modelling framework called ECLUST to test this hypothesis, and evaluate its performance through extensive simulations.\n\nWith ECLUST, we found improved prediction and variable selection performance compared to methods that do not consider the environment in the clustering step, or to methods that use the original data as features. We further illustrate this modelling framework through the analysis of three data sets from very different fields, each with high dimensional data, a binary exposure, and a phenotype of interest. Our method is available in the eclust CRAN package.

genomics↗