Send to

Choose Destination
See comment in PubMed Commons below
J Pharm Biomed Anal. 2013 Oct;84:5-13. doi: 10.1016/j.jpba.2013.05.040. Epub 2013 May 31.

Differentiation of Pueraria lobata and Pueraria thomsonii using partial least square discriminant analysis (PLS-DA).

Author information

Faculty of Pharmacy, The University of Sydney, NSW 2006, Australia.


The aims of the study were to differentiate Pueraria lobata from its related species Pueraria thomsonii and to examine the raw herbal material used in manufacturing kudzu root granules using partial least square discriminant analysis (PLS-DA). Sixty-four raw materials of P. lobata and P. thomsonii and kudzu root-labelled granules were analysed by ultra performance liquid chromatography. To differentiate P. lobata from P. thomsonii, PLS-DA models using the variables selected from the entire chromatograms, genetic algorithm (GA), successive projection algorithm (SPA), puerarin alone and six selected peaks were employed. The models constructed by GA and SPA demonstrated superior classification ability and lower model's complexity as compared to the model based on the entire chromatographic matrix, whilst the model constructed by the six selected peaks was comparable to the entire chromatographic model. The model established by puerarin alone showed inferior classification ability. In addition, the PLS-DA models constructed by the entire chromatographic matrix, GA, SPA and the six selected peaks showed that four brands out of seventeen granules were mislabelled as P. lobata. In conclusion, PLS-DA is a promising procedure for differentiating Pueraria species and determining raw material used in commercial products.


Kudzu root; Multivariate analysis; Partial least square discriminant analysis; Pueraria lobata; Pueraria thomsonii; Ultra performance liquid chromatography

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Elsevier Science
    Loading ...
    Support Center