CONSTRUCTION OF CONFIDENCE-INTERVALS FOR QUANTITATIVE TRAIT LOCI LOCATION IN F2 POPULATIONS

This paper is concerned with the construction of a confidence interval for the quantitative trait loci (QTL) location parameter in an F-2 po pulation. As usual for QTLs having small effects, this method is devel oped in the local asymptotic framework. A maximum likelihood ratio tes t is built up from statistics whose asymptotic distribution under the null hypothesis does not depend on the nuisance parameters. Simulation s show that our confidence interval gives the correct probability of c ontaining the true location of the QTL. An example using experimental data from an F-2 progeny of maize is presented to illustrate this meth odology.