Discovering Transgenic Elite Events: Using Information from Early Screening Trials for Improving Experimental Design

The transgenic improvement of a commercial crop requires a suitable transgene as well as a suitable insertion location in the crop's genome. An elite event is a successful insertion of an active transgene, and many events need to be screened to detect such elite events. This is expensive, hence optimization is needed. The screening of events in a commercial crop is usually preceded by the screening of genes in an easily transformable test plant, either a non-commercial variety of the crop or another species. This first phase mainly aims at discovering active genes. The results from the first phase can also be utilized to predict the probability of finding elite events with similar or larger effects in the commercial crop, making use of estimated variance components in the mixed model framework. The theoretical relationship between these variance components and the probability to discover in the second phase an event with an effect of a predefined magnitude is studied. The consequences of having to estimate the variance components from a first phase experiment are studied through simulations. These show that, in order to be useful to assist in the design of the second phase, the first phase should involve a large number of events, typically more than 30. The number of events is much more important than the number of plants within an event. The approach is applied to real data obtained in a commercial screening facility.