Submitted to: Crop Science
Publication Type: Peer reviewed journal
Publication Acceptance Date: 8/17/2007
Publication Date: 1/1/2008
Citation: Casa, A., Pressoir, G., Brown, P., Mitchell, S., Rooney, W., Tuinstra, M., Franks, C.D., Kresovich, S. 2008. Community resources and strategies for association mapping in sorghum. Crop Science. 48(1):30-40. Interpretive Summary: Recently, the association mapping approach has gained interest as a powerful means of identifying genes in several plant species. In order to utilize this technology, one must assemble a large set of diverse germplasm and characterize how these lines are related to one another. In this study, we assembled a set of 377 extremely diverse and historically important sorghum lines and then used molecular markers to gain an understanding of the genetic relationships among all of these distinct types. The groupings based on the molecular markers agreed quite well with what is known about the origin and appearance of the lines, and initial testing of our estimates of genetic structure show that the model we propose here should be quite effective for association mapping studies.
Technical Abstract: Association mapping is a powerful strategy for identifying genes underlying quantitative traits in plants. We have assembled and characterized genetic and phenotypic diversity of a sorghum panel suitable for association mapping, comprised of 377 accessions representing all major cultivated races (tropical lines from diverse geographic and climatic regions), and important U.S. breeding lines and their progenitors. Accessions were phenotyped for eight traits and levels of population structure and familial relatedness were assessed with 47 simple sequence repeat (SSR) loci. The panel exhibited substantial morphological variation and little genotypic differentiation was observed between the converted tropical and breeding lines. The phenotypic and genotypic data were used to evaluate the performance of several association models in controlling for spurious associations. Our analysis indicated that association models that accounted for both population structure and kinship performed better than those that did not. In addition, we found that the optimal number of subpopulations used to correct for population structure was trait-dependent. Although augmentation of the genotypic data with additional SSR loci may be necessary, the association models, genotypic data and germplasm panel described here provide a starting point for sorghum researchers to begin association studies of traits and markers/candidate genes of interest.