Principal component regression monitoring

The main goal of this chapter is to present the theoretical background of some basic chemometric methods as a tool for the assessment of surface water quality described by numerous chemical and physicochemical parameters. As a case study, long-term monitoring results from the watershed of the Struma River, Bulgaria, are used to illustrate the options offered by multivariate statistical methods such as CA, principal components analysis, principal components regression (models of source apportionment), and Kohonen s SOMs. 

For mixtures of several nucleotides and/or nucleosides, however, spectroscopic analyses become difficult because UV spectra contain only a small number of absorption bands and have a very similar shape. In these cases, it is necessary to evaluate precisely and in detail the minor differences in the spectra, such as the position of the shoulders and the relative intensities of the various bands, which are often sufficient for multicomponent analysis. Two approaches, the so-called principal component regression (PGR) and partial least-squares (PLS) regression analyses, with Fourier preprocessing of UV spectra, have been developed. These methods can monitor continuously entire spectral regions, and they are generally reliable and accmate. They have allowed the composition of 14 unknown mixtures of dAMP, dCMP, dGMP, and dTMP to be analyzed, with average per cent errors of 0.60-1.85%. The presence of a protein... 

---