Abstract
The applicability of near infrared (NIR) spectroscopy combined with chemometrics was examined to develop fast, low-cost and non-destructive spectroscopic methods for classification of transgenic maize plants. The transgenic maize plants containing both cry1Ab/cry2Aj-G10evo proteins and their non-transgenic parent were measured in the NIR diffuse reflectance mode with the spectral range of 700–1900 nm. Three variable selection algorithms, including weighted regression coefficients, principal component analysis -loadings and second derivatives were used to extract sensitive wavelengths that contributed the most discrimination information for these genotypes. Five classification methods, including K-nearest neighbor, Soft Independent Modeling of Class Analogy, Naive Bayes Classifier, Extreme Learning Machine (ELM) and Radial Basis Function Neural Network were used to build discrimination models based on the preprocessed full spectra and sensitive wavelengths. The results demonstrated that ELM had the best performance of all methods, even though the model’s recognition ability decreased as the variables in the training of neural networks were reduced by using only the sensitive wavelengths. The ELM model calculated on the calibration set showed classification rates of 100% based on the full spectrum and 90.83% based on sensitive wavelengths. The NIR spectroscopy combined with chemometrics offers a powerful tool for evaluating large number of samples from maize hybrid performance trials and breeding programs.
Highlights
Plant breeding uses molecular biology to produce new crop varieties or lines with desirable properties by using techniques to select and introduce genetic modifications and desirable traits into plants (Liu et al, 2015; Yadav et al, 2015; Yang et al, 2017)
The applicability of near infrared (NIR) spectroscopy combined with chemometrics was examined to develop fast, low-cost and non-destructive spectroscopic methods for classification of transgenic maize plants
Five classification methods, including K-nearest neighbor, Soft Independent Modeling of Class Analogy, Naive Bayes Classifier, Extreme Learning Machine (ELM) and Radial Basis Function Neural Network were used to build discrimination models based on the preprocessed full spectra and sensitive wavelengths
Summary
Plant breeding uses molecular biology to produce new crop varieties or lines with desirable properties by using techniques to select and introduce genetic modifications and desirable traits into plants (Liu et al, 2015; Yadav et al, 2015; Yang et al, 2017). One major technique of plant breeding is selection, the process of selectively propagating plants with desirable traits and eliminating those with less desirable traits (Schart et al, 2016) This requires plant breeders to screen large populations of crops for individuals that possess the characteristics of interest. There are various molecular methodologies for plant breeding, such as polymerase chain reaction (PCR) (Taverniers et al, 2004), enzyme linked immunosorbent assays (Kamle et al, 2011) and microarrays (Xu et al, 2005). These DNA- and protein-based methods for identification of transgenic plants are time consuming and costly when studying large numbers of samples, and unsuitable for on-line application. A method for the selection of transgenic samples after
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
