Journal
LWT-FOOD SCIENCE AND TECHNOLOGY
Volume 72, Issue -, Pages 63-70Publisher
ELSEVIER
DOI: 10.1016/j.lwt.2016.04.032
Keywords
Partial least square; Non-destructive; Calibration; Validation
Categories
Funding
- Professional Development Program of the Agricultural Research Council (ARC) of South Africa through ARC-VOP
- National Research Foundation (NRF) of South Africa
- University of KwaZulu-Natal
Ask authors/readers for more resources
Near-infrared spectroscopy (NIRS) is an alternative analytical method that can be used to quantify protein content in sweetpotato. It is relatively cheaper and efficient than other methods. This study was conducted to develop NIRS-based models for quantifying protein content of sweetpotato for selection or wide-area production of recommended varieties. A pool of 104 sweetpotato varieties were sampled and roots scanned using NIR spectrometer. Calibration models were developed by subjecting spectral and reference datasets to partial least squares regression. Several pre-processing methods were investigated. Models that yielded the highest coefficient of determination (R-2), residual predictive deviation (RPD) and lowest root mean square error of calibration (RMSEC) and prediction (RMSEP) were selected. Optimal model performances were obtained using second derivative pre-processing, showing the highest values of R-v(2), RMSEP and RPDv of 0.98, 0.29, and 4.0, respectively. The regression analysis indicated that informative NIR bands for quantifying protein content of sweetpotatoes ranged between 1600 and 2200 nm. The results demonstrated that NIRS is capable of predicting protein content on sweetpotatoes, rapidly and accurately. Therefore, the NIRS model developed in this study may help to quantify protein composition of sweetpotato for rapid screening of germplasm in breeding programs with high throughput and accuracy. (C) 2016 Elsevier Ltd. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available