Combining qualitative and quantitative trait data in classification of gene bank accessions

Proper classification and establishment of relationships between and within genebank accessions are very important tools for better utilization of germplasm in plant breeding. Plant morphological characteristics are often used to generate a dissimilarity matrix which serves as input to the cluster analysis. Accessions are generally described by a mixture of both qualitative and quantitative information which complicates the analysis. A wide range of distance measures is now available but only a few of them combine qualitative and quantitative data. The objective of this research was to evaluate proposed distance measures on the bases of criteria proposed originally for evaluation of different clustering algorithms.