dataset = SalmonellaTA98 begin description Data on Ames Salmonella reverse mutagenicity assay. The original authors Margolin et al. (1981) proposed the model lambda(x) = T{1 - exp(-a-bx)} exp(-cx) where x=dose level, b=mutagenic effects of dose and c=the toxic effect, T=number of microbes placed on each plate (unknown). This model can be fitted as a poisson log-linear model: log(lambda(x)) = a* + b log(x+h) - cx with c=constant set at the small dose level different from zero (10) There are replicate counts with excess variation. From: Breslow N.E. (1984), Extra-Poisson Variation in Log-linear Models, ApplStat, pp. 38-44. end description begin variables Col 0 = dose = dose levels of quinoline Col 1 = log[dose+10] = (log (+ dose 10)) Col 2 = salm = numbers of revertant colonies of TA98 Salmonella observed on each of three replicate plates testes at each of six dose levels of quinoline end variables begin data ( 0 2.30259 15 0 2.30259 21 0 2.30259 29 10 2.99573 16 10 2.99573 18 10 2.99573 21 33 3.7612 16 33 3.7612 26 33 3.7612 33 100 4.70048 27 100 4.70048 41 100 4.70048 60 333 5.83773 33 333 5.83773 38 333 5.83773 41 1000 6.91771 20 1000 6.91771 27 1000 6.91771 42 )