Estimates of the relationships between quantitative traits and molecular markers by means of Genetic Classifier Systems

Linear models are currently used to identify individuai loci responsible for the expression of quantitative traits and to predict phenotypic values from molecular profiles. Recent development of molecular technology and the consequent availability of a large number of markers can impose severe limitations on the use of linear models. The increasing complexity of genetic information requires, for this reason, new analytical tools. Among others, Genetic Classifier Systems can be a promising approach, being adaptive algorithms which deal with complex problems in an easy but powerful way. A simple Genetic Classifier System is presented and applied to a data set of reduced dimension. The results are compared with those obtained by application of a logistic model.