Genome-wide prediction and association mapping of potato common scab with historical data

Abstract Common scab, caused by Streptomyces spp., is worldwide an important skin disease of potato, capable of significant reductions in marketable value. Resistant varieties developed through phenotypic selection have been the most effective strategy thus far. Previous studies on the genetics of resistance have identified only minor-effect QTLs. In the current study, we explored the value of historical data for genetic analysis, derived from 52 sources. Based on partial replication of the 3500+ varieties, generalized (entry-mean) heritability was estimated at 0.66. For a subset of 292 varieties with genome-wide markers, the genomic (narrow-sense) heritability was only 0.10. The historical data was combined with a contemporary US dataset spanning 6 environments and 416 varieties. Genome-wide association studies identified four QTLs, which together explained 7.3% of the variation. The accuracy of genomic-estimated breeding values for marker-assisted selection was higher for the contemporary US chip group ( r = 0.71) than for European varieties from the historical dataset ( r = 0.56). This difference can be explained by a larger population size, higher genomic heritability (0.24), and higher relatedness. Using cross-validation to simulate marker-based selection, the US chip group was able to predict the European group with median accuracy 0.48, but vice versa the accuracy was only 0.29. This study has illustrated the potential for leveraging historical data for genomics-assisted breeding, but genetic gain for potato common scab resistance continues to be limited by low heritability and high polygenicity.