Submitted to: PLoS One
Publication Type: Peer reviewed journal
Publication Acceptance Date: 12/2/2010
Publication Date: 1/27/2011
Citation: Becker, A., Chao, D., Zhang, X., Borevitz, J., Salt, D.E., Baxter, I.R. 2011. Bulk segregant analysis using single nucleotide polymorphism microarrays. PLoS One. 6(1). Available: http://www.plosone.org/article/fetchObjectAttachment.action;jsessionid=D50AD35EFC46B269B0DC1E4ECB99A37B.ambra02?uri=info%3Adoi%2F10.1371%2Fjournal.pone.0015993&representation=PDF. Interpretive Summary: In order to understand all of life, it is necessary to identify the genes underlying all facets of an organism. The process of mapping to a gene has historically been a time and resource intensive endeavor. One of the limiting steps was the identification of DNA differences that can be used for mapping(markers) between two lines that have differences in a given trait. Significant advances in sequencing and microarray technologies have enabled the creation of silicon arrays with hundreds of thousands of features for assaying single DNA base changes Single Nucleotide Polymorphisms (SNP). For any given pair of crop lines, the arrays will have tens of thousands of features that can be used as markers. In this paper, we show that a SNP array designed for the model plant Arabidopsis can be used for several established mapping techniques with improved speed and cost. Large sequencing resources are available for many crop plants which would allow these approaches to be used in economically important crops, such as maize, soybean and cotton. These resources will enable plant breeders and producers to make rapid strides in crop improvement.
Technical Abstract: Bulk segregant analysis using microarrays, and extreme array mapping have recently been used to rapidly identify genomic regions associated with phenotypes in multiple species. These experiments, however require the identification of single feature polymorphisms between the cross parents for each new combination of genotypes, which raises the cost of experiments. The availability of the genomic polymorphism data in Arabidopsis thaliana, coupled with the efficient designs of Single Nucleotide Polymorphism (SNP) genotyping arrays removes the requirement for SFP detection and lowers the per array cost, thereby lowering the overall cost per experiment. To demonstrate that these approaches would be functional on SNP arrays and determine confidence intervals, we analyzed hybridizations of natural accessions to the Arabidopsis ATSNPTILE array and simulated BSA or XAM given a variety of gene models, populations, and bulk selection parameters. Our results show a striking degree of correlation between the genotyping output of both methods, which suggests that the benefit of SFP genotyping in context of BSA can be had with the cheaper, more efficient SNP arrays. As a final proof of concept, we hybridized the DNA from bulks of an F2 mapping population of a Sulfur and Selenium ionomics mutant to both the Arabidopsis ATTILE1R and ATSNPTILE arrays, which produced almost identical results. We have produced R scripts that prompt the user for the required parameters and perform the BSA analysis using the ATSNPTILE1 array and have provided them as supplemental data files.