IBD-based QTL detection In cereal breeding programs

摘要

Mapping quantitative trait loci in crop plants is usually based on bi-parental populations. Although high detection power can be achieved with such designs, the inferences that can be made are limited to the two parental lines, often leading to hardlyexploitable results. Besides, the cost of specific trials to perform QTL mapping is often a limitation for routinely map QTL. On the other hand, conventional breeding programs use a lot of crosses with many different parents. The resulting selfed progeny represents a real picture of the improved material, on which many quantitative trait notations are available because of their use during the selection process. The development of statistical techniques able to map QTL in any kind of inbred populations,like breeding populations, is challenging, mainly to be widely used by-breeders in conventional breeding programs. We present here a two step variance component technique based on the calculation of the Identity-By-Descent probabilities to routinely mapQTL in any kind of inbred pedigree. The performance of the QTL mapping methodology is assessed using simulated data which mimic an real recurrent pedigree selection. The capacity of the technique to accurately estimate parameters is investigated for a wide range of scenarios.