ARS | Publication request: Population structure and linkage disequilibrium in oat (Avena sativa L.): implications for genome-wide association studies

Submitted to: Theoretical and Applied Genetics
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: October 11, 2010
Publication Date: November 2, 2010
Citation: Newell, M.A., Cook, D., Tinker, N.A., Jannink, J. 2010. Population structure and linkage disequilibrium in oat (Avena sativa L.): implications for genome-wide association studies. Theoretical and Applied Genetics. 122:623-632.

Interpretive Summary: High density, high-throughput markers have recently become available in oat. These markers should enable association mapping of genes affecting complex traits. The study reported in this manuscript begins to answer questions concerning how best to implement association mapping in oat and what mapping precision it may provide. A first question is that of identifying population structure within oat, that is, identifying sets of lines more strongly related to each other than to the rest of the lines. These sets define appropriate materials and inference space for an analysis. In total, 1205 lines and 402 markers were used to explore population structure. For the lines used in this study, relatively weak population structure exists. To explore decay of co-inheritance of genes (that is, linkage disequilibrium, or LD), map distances of 2,225 linked marker pairs were compared to their LD (estimated as r2). Results showed that LD between linked markers decayed rapidly to r2 = 0.2 for marker pairs with a map distance of 1.0 centi-Morgan (cM). For association mapping, we suggest a minimum of one marker every cM, but higher densities of markers should increase marker–QTL association and therefore detection power. Additionally, it was found that LD was relatively consistent across oat sub-populations. These findings suggest that association mapping in oat can include germplasm with diverse origins and backgrounds. The results from this research demonstrate the feasibility of association mapping and related analyses in oat.

Technical Abstract: The level of population structure and the extent of linkage disequilibrium (LD) can have large impacts on the power, resolution, and design of genome-wide association studies (GWAS) in plants. Until recently, the topics of LD and population structure have not been explored in oat due to the lack of a high throughput, high-density marker system. The objectives of this research were to survey the level of population structure and the extent of LD in oat germplasm and determine their implications for GWAS. In total, 1205 lines and 402 diversity array technology (DArT) markers were used to explore population structure. Principal component analysis and model-based cluster analysis of these data indicated that, for the lines used in this study, relatively weak population structure exists. To explore LD decay, map distances of 2,225 linked DArT marker pairs were compared to LD (estimated as r2). Results showed that LD between linked markers decayed rapidly to r2 = 0.2 for marker pairs with a map distance of 1.0 centi-Morgan (cM). For GWAS, we suggest a minimum of one marker every cM, but higher densities of markers should increase marker–QTL association and therefore detection power. Additionally, it was found that LD was relatively consistent across the majority of germplasm clusters. These findings suggest that GWAS in oat can include germplasm with diverse origins and backgrounds. The results from this research demonstrate the feasibility of GWAS and related analyses in oat.