NIH Reporter #1R21ES039048-01

Identifying PFAS-Induced Disruption of Placental Imprinting and Neurodevelopment Through Multi-Omic Negative Control Calibration

The project investigates how prenatal PFAS exposure disrupts placental genomic imprinting and affects neurodevelopment, using multi-omic data and advanced statistical methods to identify causal mediation pathways and develop a causal inference R package.

Description

This research aims to characterize how PFAS disrupts placental parent-of-origin isoform expression using comprehensive birth cohort data and develop novel statistical methods to rigorously test mediation pathways linking environmental exposures to neurodevelopmental outcomes. The project leverages two birth cohorts with genotypes, maternal PFAS measurements, placental multi-omics data, and neurodevelopmental assessments. It employs negative binomial generalized linear mixed models for expression data and Beta regression models for DNA methylation data to identify PFAS-associated changes. It also develops and validates SCENIC, an R package implementing causal inference approaches using Generative Adversarial Networks to address unmeasured confounding. The outcomes include identifying significant genes and isoforms affected by PFAS, validating mediation pathways, and providing a causal inference R package. This work will inform understanding of PFAS effects on genomic imprinting and guide targeted interventions during critical developmental windows.

View original record