Key points are not available for this paper at this time.
The variation in amino acid sequence, in a set of bradykinin potentiating pentapeptides, is described by three variables per amino acid position. The variables were derived from a principal components analysis of a property matrix for the 20 coded amino acids. The resulting structure descriptor matrix describes the observed activity of the peptides to 97% by means of a multivariate partial least squares (PLS) model. It is demonstrated that this quantitative structure-activity relationship (QSAR) can be used to predict the activity of new peptide analogs.
Hellberg et al. (1986) studied this question.