Portable Sensors: Testing the Next Generation of Quality Assurance Devices for the Soybean Industry

1st Place Ohio Valley Institute of Food Technologists Undergraduate Research Competition2nd Place CFAES Undergraduate Research Forum3rd Place IFT17 Undergraduate Research Competition (Within the top 3 undergraduate research projects of the international Institute of Food Technologists organization)Soybean varieties have been genetically modified for increased oleic acid content to achieve a healthier and more stable oil product. With these new genetically modified organisms (GMO), it is economically important that the oil and protein content of the soybean are not reduced and that rapid methods of differentiating these new varieties and their components are found. The use of infrared (IR) spectroscopy, a form of vibrational spectroscopy, is an attractive method for the food industry due to its ability to rapidly qualify and quantify characteristics of many foods. The objective was to develop a rapid non-targeted screening approach to authenticate GMO high oleic versus non-GMO conventional soybeans by combining IR spectroscopy with pattern recognition analysis. Soybean samples (n=60) were kindly provided by DuPont Pioneer including equal numbers of GMO high oleic and non-GMO conventional varieties. Soybeans were homogenized and their composition was characterized by reference methods for fat analysis (Soxhlet, AOAC #945.16), protein analysis (Dumas, ICC Standard No. 167), and fatty acid profile (Gas Chromatography, AOAC #996.06). Spectra was collected with a portable Fourier transform infrared (FTIR) spectrometer and analyzed by Soft Independent Modeling of Class Analogy (SIMCA) and Partial Least Square Regression (PLSR) to develop classification and quantitative (fat and protein) algorithms. The GMO soybeans contained higher oleic acid content (54%), lower levels of polyunsaturated (50%) and saturated fats (4.3%), and similar levels of protein (~38%) and fat (~16%) content compared to their non-GMO counterpart. Pattern recognition analysis showed there was significant difference between the IR spectra of the GMO and non-GMO soybeans with a class distance of 6.4 and that the genetic modification had made a phenotypical difference in the chemical profile. Regression analysis associated the IR signal with both fat and protein components. Regression models were generated for non-destructive and rapid determination of oil (Rcv=0.98 and SECV=0.19) and protein (Rcv=0.98 and SECV=0.06) levels. The results showed that portable spectroscopy devices have potential to be used for a rapid, in-field, and non-destructive method to identify different varieties of soybeans and screen for specific traits, making this a great alternative to time-consuming analytical methods.A three-year embargo was granted for this item.Academic Major: Food Science and Technolog