Detection of Roundup Ready™ Soybeans by Near-Infrared Spectroscopy

Abstract

Identification and segregation of genetically modified (GMO) grains challenges the U.S. grain handling system to find a rapid and inexpensive test to distinguish GMO grain from non-GMO grain in inbound deliveries. In this study, spectra from Infratec 1220 series Whole Grain Analyzers of Roundup Ready™ and conventional soybeans were analyzed using Partial Least Squares (PLS), Locally Weighted Regression (LWR), and Artificial Neural Networks (ANN) models. Locally Weighted Regression using a database of approximately 8,000 samples, provided the most accurate classification model (93% accuracy), while ANN and PLS methods provided classification accuracies of 88% and 78%, respectively.