New Strategy for Profiling Flavonoids by an Automated Data Processing Tool: FlavonQ

Food Composition and Methods Development Laboratory, Beltsville Human Nutrition Research Center, Agricultural Research Service, U.S. Department of Agriculture, Building‐161, BARC‐East, 10300 Baltimore Avenue, Beltsville, Maryland 20705, United States Flavonoids are well‐known for their health benefits and can be found in nearly every plant. There are more than 5,000 known flavonoids existing in foods. Profiling flavonoids in natural products poses great challenges due to the diversity of flavonoids, the lack of commercially available standards, and the complexity of plant matrices. The increasingly popular use of ultra high‐performance liquid chromatography‐high resolution accurate mass‐mass spectrometry (UHPLC‐HRAM‐MS) for the analysis of flavonoids has provided more definitive information but also vastly increased amounts of data. Thus, mining of the UHPLC‐HRAM‐MS data is a very daunting, labor‐intensive, and expertise‐dependent process. An automated data processing tool, FlavonQ, was developed that can transfer field‐acquired expertise into data analysis and facilitate flavonoid research. FlavonQ is an “expert system” designed for automated data analysis of flavone/flavonol glycosides, two important subclasses of flavonoids. FlavonQ is capable of data format conversion, peak detection, flavonoid peaks extraction, flavonoid identification, and production of quantitative results. A new strategy was proposed in this study for tentative identification and quantitation of flavonoids using UHPLC HRAM‐MS n and FlavonQ. The flavonoid chromatographic peaks were firstly extracted from DAD chromatograms by FlavonQ based on their characteristic UV absorbance, filtered by similarity analysis, and then putatively identified and quantified. This approach was applied to the analysis of flavonoids in different plants with minimal user input. The data analysis process by FlavonQ was less than 1 minute per sample and both the quantitative and qualitative goals were achieved. Manual verification indicated that over 90% of flavonoids in each plant were tentatively identified and quantified correctly by the approach. The program is designed in a modular manner and allows substitution or addition of supplementary processing steps for the analysis of other classes of compounds. With FlavonQ, the time needed to perform flavonoid data analysis can be significantly reduced (hours with human verification) as compared to days or weeks needed with manual data‐mining. This project developed an expert system that used the latest chromatographic and MS technology to systematically determine flavonoids in plant materials. The composition information accrued by the FlavonQ is the infrastructure necessary to evaluate the health effect of these plant compounds, which can ultimately be used to establish dietary recommendations. Support or Funding Information This research is supported by the Agricultural Research Service of the U.S. Department of Agriculture, an Interagency Agreement with the Office of Dietary Supplements at the National Institutes of Health