|MYLES, SEAN - Nova Scotia Agricultural College|
|REISCH, BRUCE - Cornell University - New York|
|LI, LIN - Harvard University|
|Buckler, Edward - Ed|
Submitted to: Molecular Breeding
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/17/2015
Publication Date: 3/2/2015
Citation: Myles, S., Brooks, S., Harriman, J.V., Reisch, B., Ramming, D.W., Owens, C.L., Li, L., Buckler Iv, E.S., Cadle Davidson, L.E. 2015. Genetic mapping in grapevine using a SNP microarray: intensity values. Molecular Breeding. 35:88.
Interpretive Summary: Genotyping microarrays are a common technology used to assay genetic variation. These arrays rely on the matching of a sample’s DNA to a sequence from a reference plant. When a sample’s DNA differs strongly from the reference DNA, the data from the array can be difficult to interpret. This is especially the case in high diversity plant species. Here we show that much of the data from a grape genotyping microarray generates unreliable data and results in poor power for genetic mapping. However, we demonstrate that we can genetically map three simple traits (i.e. color, flower sex and resistance to powdery mildew) in the grape by using raw fluorescence data instead of software-generated sequence predictions. Our novel genetic mapping method is easy to compute and can be used in any mapping population for which genotyping microarray data have been collected. We suggest that genotyping microarrays are often unsuitable for high diversity species, which includes many of the plants we rely on for food. We recommend that future studies of this nature make use of DNA sequencing technologies rather genotyping microarrays.
Technical Abstract: Genotyping microarrays are widely used for genome wide association studies, but in high-diversity organisms, the quality of SNP calls can be diminished by genetic variation near the assayed nucleotide. To address this limitation in grapevine, we developed a simple heuristic that uses hybridization intensity to genetically map phenotypes without the need to distinguish between polymorphic states. We applied this approach to the mapping of three previously localized, qualitative traits – color, flower sex, and powdery mildew resistance – and confirmed that intensity values outperform SNP calls in all cases. Further, because per sample cost is a major limitation to the adoption of genotyping microarrays in applied genetic research and plant breeding, we tested how many samples were required to map a Mendelian trait and found that we could discover the causal locus with as few as 12 samples. For high diversity species for which genotyping arrays are available or under development, our findings provide a powerful and cost-effective approach to identify marker-trait associations when faced with poor SNP quality.