NEIGHBOUR MAPPING: AN ODD APPROACH TO LINKAGE ANALYSIS

John Innes Centre, Norwich Research Park, Colney, Norwich, NR4 7UH, UK

The order of markers in a linkage map is usually decided upon after searching among possible orders for the one which, given the data, best fits the appropriate theory. Any such searching procedure is computationally intensive, because, for n markers there are n!/2 possible linear orders. A method related to cluster analysis has been developed. This method, which owes much to the neighbour joining method of Saitou and Nei, defines an order given an input data set, and can be used to determine preliminary orders based on whole data sets where marker distribution is not ideal. Additional information can be derived from linkage data treated in similar ways and this will be presented in relation to crossover distribution and scoring errors.