AMaCAID: a useful tool for Accurate Marker Choice for Accession Identification and Discrimination

AMaCAID is an R program designed to analyse multilocus genotypic patterns in large samples. It allows (i) the computation of the number and frequency of the different multilocus patterns available in a molecular data set and (ii) the analysis of discriminatory power of each combination of k markers among n available. It thus enables the identification of the minimum number of markers required to distinguish all the observed genotypes and the subset of markers that maximize the number of distinct genotypes. AMaCAID can be used with any kind of molecular markers, on data sets mixing different kinds of markers, but also on qualitative characters like morphological or taxonomic traits. AMaCAID has been built primarily to select subsets of markers for identifying accessions and monitoring their genetic stability during regeneration cycles in an ex situ genebank. It can, however, also be used to screen any kind of data set that characterizes a set of individuals or species (e.g. taxonomic or phylogenetic studies) for discrimination purposes. The size of the assayed sample has no limitation, but the program only performs computations on all combinations of markers when there are less than 25 markers. For larger number of markers/characters, it is possible to ask AMaCAID to screen a large but limited number of combinations of markers. We apply AMaCAID to three data sets involving either molecular or taxonomic data and give some results on the computing time of the program with respect to the size of the data set.