Cyclic genotyping strategies. III: A comparison of predictive methods for group genotyping

This paper provides an investigation into some of the key practical issues for minimizing the cost of DNA testing. Previous studies focused on maximizing the utility of genotyping by prioritizing individuals for genotyping. For logistical reasons, individuals may have to be genotyped in groups rather than individually, and the best group to genotype is expected to differ from the same-sized group chosen when individuals are genotyped sequentially. In a calibration step, simulated populations and full knowledge of genotypes were used to discover the best group(s) to genotype. The characteristics of these groups were then targeted in an optimization step, using normally available information for group formation in targeted populations. Contrasts were made among predictive indices for: (i) individuals, with genotyping between each individual; (ii) individuals, with genotyping occurring group-at-a-time; and (iii) groups, using group variables as criteria. The results of this investigation allow the determination of the value of moving from individual to group genotyping, reveal the favourable attributes of individuals for group formation, and lead to methods to form groups for genotyping. The approach used has applications in reducing genotyping costs in both experimental and commercial populations for both quantitative trait loci (QTL) detection and monitoring.