Bayesian Hierarchical Repeated Measures Models to Estimate the ED50 of Known Teratogens in Sea Urchin Eggs

• Activity Number: 353
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Martiniano Flores*+ and Robert E. Weiss and Michael D. Collins
• Companies: University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles
• Keywords: Growth Curve ; Random Effects ; Teratology ; Emax

Abstract

Different doses of known human teratogens were administered to flasks containing fertilized sea urchin eggs (Strongylocentrotus purpuratus). Samples of size 60 to 100 eggs were taken, and the number of abnormal phenotypes was recorded. A key goal is to estimate the ED50, the dosage required to cause abnormal phenotypes in 50% of the sample for each teratogen. Increasing dosage increases the proportion of abnormal phenotypes, but the rate of increase is dependent on the time at which the chemical is administered, providing insight into the biological mechanism of teratogenicity. We propose several Bayesian repeated measures random effects models for the proportion of abnormal phenotypes as a function of dosage and time at which the chemical was administered, and compare the models on how well they fit the data and how they estimate the time post-fertilization where the eggs are most susceptible to the effects of chemical.