High-Performance Thin-Layer Chromatography (HPTLC) is an essential tool used extensively in the pharmaceutical industry, from general pharmaceutical analysis to identifying herbal medicines under regulated environments. Australian HPTLC Association members have been working on numerous methods from quick HPTLC screening of native super foods to comprehensive fingerprinting incorporating intensive data for automated identification and quick “in analysis” sample preparation of difficult samples used in pharmaceutical preparations.
Presentation 1: Authentication and quality control of Western Australian honeys using HPTLC
Md Khairul Islam, Division of Pharmacy, School of Allied Health, University of Western Australia
As a highly priced natural product, honey is appreciated for its nutritional value and its medicinal properties. Quality control is important but also challenging, especially for honeys collected from botanically rich and diverse areas. In this session, Md Khairul Islam, Lecturer at the University of Western Australia, explores the use of HPTLC for the authentication and quality control of honey using methods for:
- The authentication of its predominant floral source via HPTLC fingerprinting of its organic extract.
- The detection and quantification of honey constituents that might contribute to its antioxidant activity via HPTLC-DPPH analysis.
- The qualitative and quantitative analysis of its major sugars by a HPTLC assay.
Presentation 2: HPTLC and machine learning – How to use machine learning with the outputs from HPTLC
Kevin Vinsen, Senior Research Fellow, International Centre for Radio Astronomy Research (ICRAR), University of Western Australia
In this session, Kevin Vinsen, Senior Research Fellow at the University of Western Australia, discusses methods of using machine learning to analyse the light spectra produced by HPTLC and TLC Scanners (190 – 900 nm). He discusses how autoencoders can be used to improve Dynamic Time Warping (DTW) for samples with many heterogeneous peaks (such as honey or herbal extracts), and how Gaussian Mixture Models can be used to cluster the samples.