Genome wide SNP data is commonly being used to track down points in the genome that are likely to relate to a certain phenotype, and so locate the underlying gene. In this way, association mapping is a bit like quantitative trait analysis, but with one big difference. Association mapping is carried out (usually) on accessions or individuals which have been separated by a very large number of generations, consequently one can tend to assume a very short linkage drop off from a biomarker to the underlying gene. Consequently, association mapping holds the promise of 'nailing' genes quickly and cheaply. Clearly, life is never that simple. While biomarkers may be associated with a trait, they are also inevitably associated with a region or phylogenetic structure, consequently there is a danger that SNPs that may appear to be important are actually spurious. This is further complicated by the possibility, that an interesting trait that arises in a certain area will be legitimately associated with the phylogenetic structure, the region and the trait!
At ARG we are collaborating with Eric Holub at Warwick HRI to develop software to analyze SNP data from Arabidopsis thaliana. The work in this area relates to work in our modelling area in which we would like to construct models of domestication in which agronomic loci are selected for leading to linkage disequilibrium. Through modelling studies of the rise of the domestication syndrome, we aim to train association mapping techniques for the identification of important loci which have been subject to various evolutionary histories and genetic control regimes.
A. thaliana is of interest to molecular archaeobotany not only because of its model organism status, but because it is a possible human commensurate, ie a weedy species that was inadvertently moved around by humans. By answering questions about the Holocene history of Arabidopsis, we will learn more about the history of an agronomically neutral species which can provide a null background to agronomically important species. We hope to use SNP data and phenotypic data to help reconstruct the molecular ecology of Arabidopsis in the UK.