Determining Amino Acids in Soybeans by Near-Infrared Spectroscopy (NIR)

Formulating feeds to meet optimal amino acids levels require developing rapid and cost-effective techniques to measure the amino acid content of ingredients.  The limited research on developing amino acid calibrations has been reported, however, the predictive calibration models were dependent on the type of grain, sample form (whole grain or ground) and specific amino acid.  The effects of calibration (regression) method and the specific NIR spectrometer prompted a study that had two objectives: 1) To develop calibration equations for estimating amino acid composition of whole soybean using five different NIR spectrometers; and 2) To analyze the effects of amino acids/protein correlation, calibration method and type of spectrometer on the predictive ability of the calibration equations.

Calibration equations for estimating the amino acid composition in whole soybeans were developed using partial least squares, artificial neural networks and support vector machine regression methods for five models of NIR spectrometers.  The spectrometer instruments in the study were ASD LabSpec Pro, Bruker Optics/Congnis QTA, DICKEY-john OmegAnalzerG, and the FOSS Intratec 1241 grain analyzer.  Findings reported were that NIR instruments were usable for research purposes and sample screening; correlation of cysteine and tryptophan had no useful correlation with spectral calibrations; correlation of methionine and serine were useful for rough sample screening; and the calibrations for the other amino acids were useful for sample screening.  Bottom line is that NIR instruments and calibrations are available for determining amino acids in whole soybeans, and the level of an individual amino acid is highly correlated to the protein contained in the soybean seed.

Kovalenko, Igor and co-workers.  2006.  Determination of amino acids composition of soybean (Glycine Max) by near-infrared spectroscopy.  J. Agric. Food Chem. 54:3485-3491.