Prediction of white cabbage (Brassica oleracea var. capitata) self-incompatibility based on neural network and discriminant analysis of complex electrophoretic patterns

Self-incompatibility (SI) of white cabbage (Brassica oleracea var. capitata), a biological mechanism triggering the ability to pollinate plant ovules, is modelled based on the statistical analysis of electrophoretic patterns of stigma extracts. The patterns obtained for 72 plants grown in three different seasons (2003–2005) have been explored with discriminant analysis and non-linear neural networks. The NN models obtained perform a flawless classification of SI and provide a useful and fast technique for SI prediction before the end of the season. The discriminant analysis turns out to be less effective (74% of good predictions), but together with sensitivity analysis of NN, it points out the important markers of the SI process (peaks 1, 3, 5, 15 and 18).