COMPUTER ANALYSIS OF DISEASE DATA TO DERIVE HYPOTHETICAL GENOTYPES FOR REACTION OF HOST VARIETIES TO PATHOGENS

Abstract

A method is described for deriving genotypes for low reaction in varieties of a host from data resulting from inoculating many host varieties with many cultures of a pathogen. A computer program utilizing Boolean algebra was used to compare and group patterns of low and high infection types produced by host variety: pathogen culture combinations. Visual inspection within each group allows derivation of the genotype. A successful trial of the system was made with data on wheat leaf rust involving 60 varieties of wheat and nine cultures of Puccinia recondita.