Research Article Prediction of Radix Astragali Immunomodulatory Effect of CD80 Expression from Chromatograms by Quantitative Pattern-Activity Relationship Michelle Chun-har Ng, 1 Tsui-yan Lau, 2 Kei Fan, 1 Qing-song Xu, 3 Josiah Poon, 4 Simon K. Poon, 4 Mary K. Lam, 5 Foo-tim Chau, 2 and Daniel Man-Yuen Sze 6 1 Department of Health Technology and Informatics, Te Hong Kong Polytechnic University, Hung Hom, Hong Kong 2 Department of Applied Biology and Chemical Technology, Te Hong Kong Polytechnic University, Hung Hom, Hong Kong 3 School of Mathematics and Statistics, Central South University, Changsha 410083, China 4 School of Information Technologies, Te University of Sydney, Lidcombe, NSW, Australia 5 Faculty of Health, University of Technology Sydney, Ultimo, NSW, Australia 6 School of Health and Biomedical Sciences, RMIT University, Melbourne, VIC, Australia Correspondence should be addressed to Foo-tim Chau; foo-tim.chau@polyu.edu.hk and Daniel Man-Yuen Sze; daniel.sze@rmit.edu.au Received 9 September 2016; Revised 15 December 2016; Accepted 15 January 2017; Published 28 February 2017 Academic Editor: Adair Santos Copyright © 2017 Michelle Chun-har Ng et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Te current use of a single chemical component as the representative quality control marker of herbal food supplement is inadequate. In this CD80-Quantitative-Pattern-Activity-Relationship (QPAR) study, we built a bioactivity predictive model that can be applicable for complex mixtures. Trough integrating the chemical fngerprinting profles of the immunomodulating herb Radix Astragali (RA) extracts, and their related biological data of immunological marker CD80 expression on dendritic cells, a chemometric model using the Elastic Net Partial Least Square (EN-PLS) algorithm was established. Te EN-PLS algorithm increased the biological predictive capability with lower value of RMSEP (11.66) and higher values of  2  (0.55) when compared to the standard PLS model. Tis CD80-QPAR platform provides a useful predictive model for unknown RA extract’s bioactivities using the chemical fngerprint inputs. Furthermore, this bioactivity prediction platform facilitates identifcation of key bioactivity-related chemical components within complex mixtures for future drug discovery and understanding of the batch-to-batch consistency for quality clinical trials. 1. Introduction A large pool of medicinal plants from Chinese herbal medicines (CHM) has a long historical clinical practice for more than 2000 years ago. However, the underlying mecha- nisms of action of the CHM remain largely unknown except the few examples of taxol [1] for anticancer, artesunate [2] for malaria treatment, and arsenic trioxide [3] for leukemia treatment. While these three herbal derived single com- pounds are responsible for the efective therapies, however, for most of the other clinically useful CHM, the mechanisms of action have been considered as that of “multicompound multitarget.” Te use of herbal formula by combining a few herbs based on the Chinese medicine theory further adds to this complexity. Tus, there exists a wide range of possible chemical compounds in each single herb or complex formula that may contribute to the clinical efcacy, but this crucial information is basically unknown at the moment. Tis lack of understanding of the active compounds and their targets in turn makes the quality control aspect of ensuring the batch- to-batch consistency of CHM difcult if not impossible. Up to now, a CHM product PHY906 which is undergoing phase 2 clinical trial and being marketed as an adjuvant to chemotherapy attempted to address the batch-to-batch Hindawi BioMed Research International Volume 2017, Article ID 3923865, 11 pages https://doi.org/10.1155/2017/3923865