Usefulness of gene information in marker-assisted recurrent selection: A simulation appraisal

Genomics and post-genomics sciences are expected to uncover most, if not all, of the quantitative trait loci (QTL) in plants. Prior knowledge of QTL locations can then be exploited in marker-assisted recurrent selection (MARS). Our objectives were to determine (i) whether prior knowledge of QTL locations is advantageous in MARS, and (ii) whether knowledge of the QTL themselves, as opposed to knowledge of markers linked to QTL, is advantageous in MARS. We simulated MARS in a maize (Zea mays L.) F₂ population. We found that when 10 QTL controlled the trait, the percentage of known QTL that maximized the response to MARS was P_Max = 100%. In contrast, P_Max was often less than 100% when 40 or 100 QTL controlled the trait and QTL effects were estimated with a population size (N = 100) typically used in MARS. This result implied it was advantageous to exploit only the QTL with large effects and ignore those with small effects, even if the locations of all QTL were known. For a trait controlled by 40 QTL, the response was up to 50% greater when P_Max = 70% of the QTL were known through markers for the QTL themselves rather than through linked markers. We conclude that having known QTL in MARS is most beneficial for traits controlled by a moderately large number of QTL (e.g., 40). We speculate that a combination of approaches would be needed to exploit information on markers for QTL themselves, markers linked to QTL, and unknown QTL.