A Bayesian approach for inferring global points of departure from transcriptomics data
Joe Reynolds, Sophie Malcomber, Andrew White
Abstract
Bayesian statistical methods allow for robust scientific inferences. Increased robustness is achieved by using prior distributions to regularise parameter estimates and by defining a model structure which accurately reflects the variance structure of the dataset of interest. We develop a Bayesian model to describe transcriptomics concentration–response data. This model is designed to infer gene-level points of departure and a method to derive a global point of departure from these thousands of gene level estimates is presented. We believe such estimates may prove useful for characterising maximum no effect concentrations for the purposes of hazard identification in next generation risk assessment.