CLASSIFICATION OF NATURAL GUMS .8. CHEMOMETRIC ASSIGNMENT OF COMMERCIAL GUM EXUDATES FROM AFRICA USING CLUSTER-ANALYSIS ON THE PROTEIN AMINO-ACID COMPOSITIONS

Cluster analysis methods based on taxi distances between pairs of entities have been developed to classify commercial gum arabic in relation to authenticated Acacia senegal for regulatory and quality control purposes. Two parameters are defined for these gums which measure the closeness of any gum specimen to A.senegal. One of the parameters is described in terms of multiple nearest neighbours yielding a number designated the 'Senegal Number', and the other is a discriminant function of taxi distances of all gums from the centroids of two pre-eminent classes. The chemical data used are the amino acid compositions of the protein moiety in the gum molecule. Using this system commercial gum arabic can be distinguished from other arabinogalactan protein gums of Combretum and Albizia origin, which are potential adulterants in terms of existing food regulations.