Application of Vis/NIR diffuse reflectance spectroscopy to the detection and identification of transgenic tomato leaf

The feasibility of Vis/NIR spectroscopy technique for rapid and non-invasive detection of transgenic tomato leaves from conventional ones was investigated by means of spectral diffuse reflectance mode. A total of 68 samples (38 transgenic ones and 30 non-transgenic ones) were used for classification. The calibration and validation results were analyzed via discriminant analysis (DA) and partial least squares (PLS) discriminant method using TQ 6. 2. 1 quantitative software. Models based on the different spectral pre-processing methods (multiplicative signal correction (MSC), first and second derivative) were compared. It was found that the classification accuracy using DA was higher than that using PLS and the best results were gained by using spectra after MSC with InGaAs detector and the classification accuracy was 89.7% (accuracy of 86.8% for transgenic samples and 93.3% for non-transgenic ones). The results show that Vis-NIR diffuse reflectance spectroscopy technique is a feasible and fast method for non-invasive detection of transgenic and non-transgenic tomato leaves.