Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


final thesis.pdf (1.44 MB)

ETD Abstract Container

Abstract Header

Development of a rapid and in-field phenotyping tool for screening protein quality in soybeans (Glycine max) using a miniature near infrared sensor

Sia, Xin Rong
http://rave.ohiolink.edu/etdc/view?acc_num=osu1574800276913103

Abstract Details

2019, Master of Science, Ohio State University, Food Science and Technology.
Soybean is an economically important crop that is a major plant-based protein source for livestock diets, with the amino acid composition of soybeans being crucial for determining the quality of livestock feed. Although protein quality monitoring is important, conventional protein and amino acid analyses typically involve laborious and lengthy processes. Unsurprisingly, soybean growers and breeders have identified time-consuming wet chemistry analytical methods as a major bottleneck in improving their breeding practices, calling for faster techniques to analyze amino acids in soybeans. For instance, classical amino acid analysis methods such as ion-exchange chromatography with ninhydrin derivatization require 60 – 120 minutes of analysis time per sample and limited selectivity due to the use of optical detectors, which cannot resolve overlapping peaks. A faster alternative is the use of portable near-infrared (NIR) spectroscopy equipment combined with chemometrics that allows for direct measurement of ground soybean and even intact soybean seeds in real-time. ¬¬Our objective was to develop and evaluate the feasibility of using a sensor-based method for in-field analysis of amino acid composition in soybeans. Twenty-two soybean samples of different cultivars and grown over a period of two years across the Midwest region were selected for analysis, in addition to nineteen soy isolates, concentrates and powders obtained via online retailers. In order to develop a reliable NIR prediction model, we first needed a reliable reference method for profiling the amino acid content of the soybeans, so propyl chloroformate derivatization (PCD) coupled to gas chromatography-mass spectrometry (GC-MS) was performed to obtain the amino acid values of soybeans. GC-MS results showed high sensitivity with a LOQ of 1.1 – 14.0 ppm depending on the type of amino acid, high selectivity, and calibration curves with good linearity (R > 0.97 for most amino acids). External validation of our method with a classical amino acid analysis that uses ion-exchange chromatography with ninhydrin derivatization showed that our method is comparable in accuracy, with a correlation of R2 = 0.98, but precision needs to be improved. The largest sources of experimental errors originated from the solid-phase extraction, derivatization, and protein hydrolysis steps. Protein hydrolysis variables that had the most influence on amino acid yield was found to be the mass of samples, hydrolysis errors, and type of oxidation inhibitor used so it is recommended that these parameters are preferentially optimized. Our method demonstrated faster run times and higher selectivity than classical methods, allowing chromatographic analysis to be completed in as little as 10 mins per sample, and co-eluting peaks were successfully resolved due to the monitoring of mass fragments. Spectral collection was done using both ground soybeans and intact soybean seeds and analyzed by partial least squares regression (PLSR) to develop calibration models for predicting total protein and critical amino acid (lysine, threonine, methionine, tryptophan, cysteine) levels in soybean. The miniature NIR device we used is the first handheld device on the market to provide a spectral scanning range of between 1350 – 2500 nm, covering most of the first overtones and combination bands. This is in contrast with other miniature devices which tend to scan at lower wavelengths and cover second overtone bands, which gives less specific chemical information about the food constituents scanned. Combining spectral information with reference amino acid values determined using the classical method allowed us to build prediction models that showed good linear correlation between spectra and amino acids (r > 0.97 for ground samples, r > 0.94 for intact seeds) with low standard error of cross-validation (1.630% for protein, 0.041 – 1.630% for amino acids). Our findings support that a miniature spectrometer combined with pattern recognition is capable of real-time monitoring of important amino acids in soybeans. We used a miniature device that employed Micro Electro Mechanical Systems (MEMS) technology, resembling the quality of a Michelson interferometer with improved band resolution. The higher sensitivity and accuracy of MEMS is superior to some other miniature NIR spectrometers on the market and allowed us to successfully characterize the amino acid profile of soybeans in as little as 15 seconds.
Luis Rodriguez-Saona (Advisor)
Rafael Jimenez-Flores (Committee Member)
Matthias Klein (Committee Member)
Analytical Chemistry; Food Science
near-infrared spectroscopy; rapid phenotyping; amino acid; amino acid analysis; GC-MS; method validation

Recommended Citations

Citations

  • Sia, X. R. (2019). Development of a rapid and in-field phenotyping tool for screening protein quality in soybeans (Glycine max) using a miniature near infrared sensor [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1574800276913103

    APA Style (7th edition)

  • Sia, Xin Rong. Development of a rapid and in-field phenotyping tool for screening protein quality in soybeans (Glycine max) using a miniature near infrared sensor. 2019. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1574800276913103.

    MLA Style (8th edition)

  • Sia, Xin Rong. "Development of a rapid and in-field phenotyping tool for screening protein quality in soybeans (Glycine max) using a miniature near infrared sensor." Master's thesis, Ohio State University, 2019. http://rave.ohiolink.edu/etdc/view?acc_num=osu1574800276913103

    Chicago Manual of Style (17th edition)

osu1574800276913103
1,121
© 2019, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.