POWER STUDIES IN THE ESTIMATION OF GENETIC-PARAMETERS AND THE LOCALIZATION OF QUANTITATIVE TRAIT LOCI FOR BACKCROSS AND DOUBLED HAPLOID POPULATIONS

A statistical model for doubled haploids and backcrosses based on the interval-mapping methodology has been used to carry out power studies to investigate the effects of different experimental designs, heritabi lities of the quantitative trait, and types of gene action, using two test statistics, the F of Fisher-Snedecor and the LOD score. The doubl ed haploid experimental design is more powerful than backcrosses while keeping actual type I errors similar to nominal ones. For the doubled haploid design, individual QTLs, showing heritabilities as low as 5% were detected in about 90% of the cases using only 250 individuals. Th e power to detect a given QTL is related to its contribution to the he ritability of the trait. For a given nominal type I error, tests using F values are more powerful than with LOD scores. It seems that more c onservative levels should be used for the LOD score in order to increa se the power and obtain type I errors similar to nominal ones.