is approximately normal. The test statistic is \Z|. The 5%-significance level can be taken to be 1.96, the 5%-cutoff level for a two-sided test for the standard normal distribution. Such a level will ensure that, on the average, only one in every 20 experiments testing a single marker will falsely claim association with the phenotype when, in fact, such association is absent.

The function "Z.BC" computes the test statistic for a backcross trial. The input is a list containing the genetic and phenotypic data from the trial. The output is the value of Z based on information in the second column of the genetic matrices. (In the simulations below we will assume that the first column represents the QTL and the second represents the marker.)

0 0

Post a comment