You are here:
Computational Integration of Human Genetic and Toxicological Data to Evaluate AOP-Specific Susceptibility
Chamberlin, J. AND H. Mortensen. Computational Integration of Human Genetic and Toxicological Data to Evaluate AOP-Specific Susceptibility. Society of Toxicology, San Antonio, TX, March 11 - 15, 2018.
This project presents a computational workflow that implements the AOP-DB to accept AOP molecular events as input, automating downstream analyses that include regulatory region identification, eQTL information for AOP-relevant tissue types, and allele and haplotype frequency information for human populations.
Susceptibility to environmental chemicals can be modulated by genetic differences. Direct estimation of the genetic contribution to variability in susceptibility to environmental chemicals is only possible in special cases where there is an observed association between exposure and effect (e.g. genotype and phenotype information). The availability of genetic and toxicological data sources makes it possible to indirectly estimate the relative contribution of genetic variability to differential human susceptibility. The purpose of this project is to develop a computational workflow that integrates genetic and toxicological resources. This approach implements the Adverse Outcome Pathway (AOPs) framework in order to integrate molecular targets associated with AOPs with functional genomic annotations and population allele frequencies. Resources include the EPA internal Adverse Outcome Database (AOP-DB), and publicly available resources, such as the AOP-wiki, Ensembl genomic annotations, expression Quantitative Trait loci identified by the GTEx consortium, and 1000 Genomes Project. With this information it is possible to formulate predictions of genetic susceptibility built upon established toxicological and genetic knowledge that are specific to an adverse outcome. The computational workflow presented here is written in R and built around the Ensembl database interfaces (REST API and biomaRt R package). It downloads, integrates, and analyzes the available data sources when an AOP is given as input. Data processing involves four steps: 1.Genetic identities of AOP key events are extracted from the AOP-DB; 2. Nearby regulatory annotations are downloaded from the Ensembl regulatory build; 3. GTEx Expression quantitative trait loci are imported for AOP-relevant tissue types; and 4. Allele and haplotype frequency information is retrieved from the 1000 Genomes Project stage 3 dataset in order to quantify the degree of genetic variation at functionally relevant loci. With ongoing AOP development, this automated workflow will allow rapid assessment of outcome specific human genetic susceptibility. This abstract does not reflect EPA Policy