Potential of descriptive linear discriminant analysis for studying clinical chemical and haematological data from persons with heterozygous sickle cell disease

Predicting the probability of helper T cell immunodominant sites through discriminant analysis

Bayesian discriminant analysis is used to predict whether or not a given protein segment will activate helper T cells. The predictor variables are drawn from the products of frequencies of amino acid residues. The model's predictive validity compares favourably with that of alternative modelling strategies, suggesting that this approach merits further investigation.