Predicting Low Dose Effects for Chemicals in High Through-Put Studies

摘要: High through-put studies commonly use automated systems with 96-well plates in which multiple chemicals are tested at doses using log-2 dose increments after a suitable incubation period. There typically (ranging from five to eleven) on each chemical, and occasionally plate replications of the dose-response studies. The target endpoint for such is LC50, but some chemicals, there may be below benchmark where no apparent adverse response relative control response. We show how an estimation approach can lead clearly interpretable results about low region data high throughput study 2189 yeast. Accurate estimates obtained by best linear unbiased predictors (BLUPs) mixed model, summarized via plots expected (assuming low-dose effect) confidence intervals providing informative summary doses. conclude that this provide valuable insights would missed if observational were only considered through lens statistical methods appropriate experimental