Send to

Choose Destination
J Chromatogr A. 2008 May 23;1192(1):81-8. doi: 10.1016/j.chroma.2008.03.052. Epub 2008 Mar 22.

Exploration of linear multivariate calibration techniques to predict the total antioxidant capacity of green tea from chromatographic fingerprints.

Author information

Department of Analytical Chemistry and Pharmaceutical Technology, Vrije Universiteit Brussel, VUB, Laarbeeklaan 103, B-1090 Brussels, Belgium.


Nowadays fingerprinting is a generally applied technique for the identification and quality assessment of herbal products. In this study it was aimed to predict a quantitative property, the antioxidant capacity of green tea, from chromatographic fingerprints. Different linear multivariate calibration techniques, commonly applied on spectral data, were explored and compared. When the chromatograms were appropriately pretreated, all tested techniques were able to predict the total antioxidant capacity with a precision comparable to that of the reference method (Trolox equivalent antioxidant capacity assay). Stepwise multiple linear regression (MLR) however is less recommended because of inadequate variable selection. Principal components regression (PCR) also seems less preferable, because large variations not correlated with the total antioxidant capacity were also included in the model. This problem does not occur with partial least squares (PLS) models. Of all tested PLS methods, orthogonal projections to latent structures (O-PLS) was preferred because of its simplicity, reproducibility, good interpretability of the compounds' contribution to the antioxidant capacity and its good predictive and describing abilities.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center