Bootstrap classification and point-based feature selection from age-staged mouse cerebellum tissues of matrix assisted laser desorption/ionization mass spectra using a fuzzy rule-building expert system

A bootstrap method for point-based detection of candidate biomarker peaks has been developed from pattern classifiers. Point-based detection methods are advantageous in comparison to peak-based methods. Peak determination and selection are problematic when spectral peaks are not baseline resolved or on a varying baseline. The benefit of point-based detection is that peaks can be globally determined from the characteristic features of the entire data set (i.e., subsets of candidate points) as opposed to the traditional method of selecting peaks from individual spectra and then combining the peak list into a data set. The point-based method is demonstrated to be more effective and efficient using a synthetic data set when compared to using Mahalanobis distance for feature selection. In addition, probabilities that characterize the uniqueness of the peaks are determined.This method was applied for detecting peaks that characterize age-specific patterns of protein expression of developing and adult mouse cerebella from matrix assisted laser desorption/ionization (MALDI) mass spectrometry (MS) data. The mice comprised three age groups: 42 adults, 19 14-day-old pups, and 16 7-day-old pups. Three sequential spectra were obtained from each tissue section to yield 126, 57 and 48 spectra for adult, 14-day-old pup, and 7-day-old pup spectra, respectively. Each spectrum comprised 71,879 mass measurements in a range of 3.5-50 kDa. A previous study revealed that 846 unique peaks were detected that were consistent for 50% of the mice in each age group (C. Laurent, D.F. Levinson, S.A. Schwartz, P.B. Harrington, S.P. Markey, R.M. Caprioli, P. Levitt, Direct profiling of the cerebellum by MALDI MS: a methodological study in postnatal and adult mouse, J. Neurosci. Res. 81 (2005) 613-621.).A fuzzy rule-building expert system (FuRES) was applied to investigate the correlation of age with features in the MS data. FuRES detected two outlier pup-14 spectra. Prediction was evaluated using 100 bootstrap samples of 2 Latin-partitions (i.e., 50:50 split between training and prediction set) of the mice. The spectra without the outliers yielded classification rates of 99.1 ± 0.1%, 90.1 ± 0.8%, and 97.0 ± 0.6% for adults, 14-day-old pups, and 7-day-old pups, respectively. At a 95% level of significance, 100 bootstrap samples disclosed 35 adult and 21 pup distinguishing peaks for separating adults from pups; and 8 14-day-old and 15 7-day-old predictive peaks for separating 14-day-old pup from 7-day-old pup spectra. A compressed matrix comprising 40,393 points that were outside the 95% confidence intervals of one of the two FuRES discriminants was evaluated and the classification improved significantly for all classes. When peaks that satisfied a quality criterion were integrated, the 55 integrated peak areas furnished significantly improved classification for all classes: the selected peak areas furnished classification rates of 100%, 97.3 ± 0.6%, and 97.4 ± 0.3% for adult, 14-day-old pups, and 7-day-old pups using 100 bootstrap Latin partitions evaluations with the predictions averaged. When the bootstrap size was increased to 1000 samples, the results were not significantly affected. The FuRES predictions were consistent with those obtained by discriminant partial least squares (DPLS) classifications.     

An emphatic orthogonal signal correction-support vector machine method for the classification of tissue sections of endometrial carcinoma by near infrared spectroscopy

A new application of emphatic orthogonal signal correction (EOSC) for baseline correction of near infrared spectra from reflectance measurements of tissue sections is introduced. EOSC was evaluated and compared with principal component orthogonal signal correction (PC-OSC) by using support vector machine (SVM) classifiers. In addition, some exemplary synthetic data sets were created to characterize EOSC coupled to SVM for classification. Orthogonal experimental design coupled with analysis of variance (ANOVA) was used to determine the significant parameters for optimization, which were the OSC method and number of components for the model. EOSC combined with the SVM gave better predictions with respect to a larger number of components and was not as susceptible to overfitting the data as the classifier built with PC-OSC data. These results were supported by simulations using synthetic data sets. EOSC is a softer signal correction approach that retains more signal variance which was exploited by the SVM. Classification rates of 93 ± 1% were obtained without orthogonal signal correction with the SVM. PC-OSC and EOSC data gave similar peak prediction accuracies of 94 ± 1%. The key advantages demonstrated by EOSC were its resistance to overfitting, fine-tuning capability or softness, and the retention of spectral features after signal correction.     

(Liquid + liquid) equilibria of (water + propionic acid + alcohol) ternary systems

(Liquid + liquid) equilibrium (LLE) data of the solubility (binodal) curves and tie-line end composition were examined for mixtures of {water (1) + propionic acid (2) + octanol or nonanol or decanol or dodecanol (3)} at T = 298.15 K and 101.3 ± 0.7 kPa. The reliability of the experimental tie-line data was confirmed by using the Othmer-Tobias correlation. The LLE data of the ternary systems were predicted by UNIFAC method. Distribution coefficients and separation factors were evaluated for the immiscibility region.     

Phase equilibria of (water + levulinic acid + dibasic esters) ternary systems

Liquid–liquid equilibrium (LLE) data of the solubility (binodal) curves and tie-line end compositions were examined for mixtures of {(water (1) + levulinic acid (2) + dimethyl succinate or dimethyl glutarate or dimethyl adipate (3)} at 298.15 K and 101.3 ± 0.7 kPa. The reliability of the experimental tie-line data was confirmed by using the Othmer–Tobias correlation. The LLE data of the ternary systems were predicted by UNIFAC method. The LLE data were correlated fairly well with UNIQUAC and NRTL models, indicating the reliability of the UNIQUAC and NRTL equations for these ternary systems. The best results were achieved with the NRTL equation, using non-randomness parameter (α = 0.3) for the correlation. Distribution coefficients and separation factors were measured to evaluate the extracting capability of the solvents.     

Liquid-liquid equilibria of (water + butyric acid + diethyl succinate or diethyl glutarate or diethyl adipate) ternary systems

Liquid-liquid equilibrium (LLE) data of the solubility (binodal) curves and tie-line end compositions were examined for mixtures of {(water (1) + butyric acid (2) + diethyl succinate or diethyl glutarate or diethyl adipate (3)} at 298.2 K and 101.3 ± 0.7 kPa. The relative mutual solubility of butyric acid is higher in the diethyl succinate or diethyl glutarate or diethyl adipate layers than water layers. The consistency of the experimental tie-lines was determined through the Othmer-Tobias correlation equation. The LLE data were correlated with NRTL model, indicating the reliability of the NRTL equations for these ternary systems. The best results were achieved with the NRTL equation, using non-randomness parameter (α = 0.3) for the correlation. Distribution coefficients and separation factors were measured to evaluate the extracting capability of the solvents.     

Classification of jet fuels by fuzzy rule-building expert systems applied to three-way data by fast gas chromatography--fast scanning quadrupole ion trap mass spectrometry

A fast method that can be used to classify unknown jet fuel types or detect possible property changes in jet fuel physical properties is of paramount interest to national defense and the airline industries. While fast gas chromatography (GC) has been used with conventional mass spectrometry (MS) to study jet fuels, fast GC was combined with fast scanning MS and used to classify jet fuels into lot numbers or origin for the first time by using fuzzy rule-building expert system (FuRES) classifiers. In the process of building classifiers, the data were pretreated with and without wavelet transformation and evaluated with respect to performance. Principal component transformation was used to compress the two-way data images prior to classification. Jet fuel samples were successfully classified with 99.8 ± 0.5% accuracy for both with and without wavelet compression. Ten bootstrapped Latin partitions were used to validate the generalized prediction accuracy. Optimized partial least squares (o-PLS) regression results were used as positively biased references for comparing the FuRES prediction results. The prediction results for the jet fuel samples obtained with these two methods were compared statistically. The projected difference resolution (PDR) method was also used to evaluate the fast GC and fast MS data. Two batches of aliquots of ten new samples were prepared and run independently 4 days apart to evaluate the robustness of the method. The only change in classification parameters was the use of polynomial retention time alignment to correct for drift that occurred during the 4-day span of the two collections. FuRES achieved perfect classifications for four models of uncompressed three-way data. This fast GC/fast MS method furnishes characteristics of high speed, accuracy, and robustness. This mode of measurement may be useful as a monitoring tool to track changes in the chemical composition of fuels that may also lead to property changes.     

Classification of edible and lampante virgin olive oil based on synchronous fluorescence and total luminescence spectroscopy

Total luminescence and synchronous fluorescence spectroscopies were tested as regards their ability to differentiate edible from lampante virgin olive oils. Total luminescence spectra were recorded by measuring the emission spectra in the range 350-720 nm at excitation wavelengths from 320 to 535 nm. The synchronous fluorescence spectra of 41 edible and 32 lampante virgin olive oils were acquired by synchronous scanning the excitation and emission monochromator maintained at an offset value of 80 nm. Classification of virgin olive oils based on their synchronous fluorescence spectra was performed by hierarchical cluster analysis and principal component analysis using the spectral range of 429-545 nm. Principal component analysis provided better discrimination between the two classes, without any classification error, while hierarchical cluster analysis allowed 97.3% correct classification. These results indicate the capability of fluorescence techniques to differentiate virgin olive oils according to their quality.     

Experimental validation of an effective carbon number-based approach for the gas chromatography–mass spectrometry quantification of ‘compounds lacking authentic standards or surrogates’

For the quantitative analysis of ‘compounds lacking authentic standards or surrogates’ (CLASS) in environmental media, we previously introduced an effective carbon number (ECN) approach to develop an empirical equation for the prediction of their response factor (RF). In this research, a series of laboratory experiments were carried out to benchmark the reliability of an ECN approach for sorbent tube/thermal desorption/gas chromatography (GC)/mass spectrometry (MS) applications. First, the ECN values were determined using external calibration data from 25 reference volatile organic compounds (VOCs) using two MS dectectors (quadrupole (Q) and time-of-flight (TOF)). Then, a certified standard mixture of 54 VOCs was analyzed by each system as a simulated unknown sample. The analytical bias, assessed in terms of percentage difference (PD) between the certified and ECN-predicted mass values, averaged 19.2 ± 16.1% (TOF-MS) and 28.2 ± 27.6% (Q-MS). The bias using a more simplified carbon number (CN)-based prediction increased considerably, yielding 53.4 ± 53.3% (TOF-MS) and 61.7 ± 81.3% (Q-MS). However, the bias obtained using the ECN-based prediction decreased significantly to yield average PD values of 9.84 ± 7.28% (TOF-MS) and 16.8 ± 8.35% (Q-MS), if the comparison was limited to 26 (out of 54) VOCs with CN ≥ 4 (i.e., 25 aromatics and hexachlorobutadiene).     

Characterization of Adulterated Milk by Two-Dimensional Infrared Correlation Spectroscopy

A new approach for discrimination of adulterated milk is reported using two-dimensional infrared (IR) correlation spectroscopy by multiway principal component analysis (MPCA) and least squares support vector machines (LS–SVM). First, the synchronous two-dimensional spectra of pure and adulterated milk were calculated. Then, MPCA was used to reduce the dimensions, extract features of two-dimensional correlation data set, and distinguish adulterated milk and pure milk. Finally, a LS-SVM model was developed using the scores of the first thirteen principal components from synchronous two-dimensional correlation spectra computed by MPCA as the input variables. The ratios of correct classification were 100% and 96.3% for calibration set and prediction set, respectively. The area under the receiver operating characteristic curves (ROC) of 0.991 for prediction set was obtained by LS–SVM. The results indicate that two-dimensional correlation infrared spectra combined with MPCA–LS–SVM may be a rapid screening technique for discrimination of adulterated milk with good accuracy.     

Approach for rapid extraction and speciation of mercury using a microtip ultrasonic probe followed by LC-ICP-MS

A fast method for mercury extraction from biological samples based on the use of HCl leaching plus different enzymatic hydrolysis (with and without mercury complexing agents), and the use of focussed ultrasounds (2-mm microtip) is here proposed. Total mercury content in several biological samples was determined by FI-ICP-MS using a carrier solution consisting of 0.1% (v/v) HCl, 0.1% (v/v) 2-mercaptoethanol, to avoid memory effect, and 0.15% (w/v) KCl. For mercury speciation a RP18 chromatographic column coupled to ICP-MS was used. A mobile phase consisting of 0.1% (v/v) formic acid, 0.1% (v/v) HFBA, 2% (v/v) methanol, and 0.02% (w/v) mM l-cysteine at pH 2.1 was used for chromatographic separation of the mercury species in the sample extracts. Extraction procedures were validated by using 50 mg of tuna fish tissue CRM-463 (2.85 ± 0.16 mg kg⁻¹ for methylmercury). The recoveries obtained were 99 ± 3% and 93 ± 1% after acid leaching (HCl 7 M) and enzymatic extraction (15 mg protease type XIV in 2.5% (v/v) 2-mercaptoethanol), respectively. The optimal sonication conditions (5 min of exposure time and 40% of ultrasound amplitude) were applied to 5 mg of CRM-463 (88 ± 5%), 5 mg of mussel tissue (81 ± 11%) and to 2 mg of zebra fish embryos (90 ± 10%) obtaining good recoveries in all cases. Methylmecury was found to be the most abundant Hg specie in all samples. The developed method is simple and rapid (5 min sample treatment); it is suitable for very small samples and does not alter the original form of the mercury species. Thus, it is of special interest in those cases in which validation of the results may often be hampered by lack of sample availability.     

Synthesis and characterization of novel elastomeric poly(D,L-lactide urethane) maleate composites for bone tissue engineering

Here, we report the synthesis and characterization of a novel 4-arm poly(lactic acid urethane)-maleate (4PLAUMA) elastomer and its composites with nano-hydroxyapatite (nHA) as potential weight-bearing composite. The 4PLAUMA/nHA ratios of the composites were 1:3, 2:5, 1:2 and 1:1. FTIR and NMR characterization showed urethane and maleate units integrated into the PLA matrix. Energy dispersion and Auger electron spectroscopy confirmed homogeneous distribution of nHA in the polymer matrix. Maximum moduli and strength of the composites of 4PLAUMA/nHA, respectively, are 1973.31 ± 298.53 MPa and 78.10 ± 3.82 MPa for compression, 3630.46 ± 528.32 MPa and 6.23 ± 1.44 MPa for tension, 1810.42 ± 86.10 MPa and 13.00 ± 0.72 for bending, and 282.46 ± 24.91 MPa and 5.20 ± 0.85 MPa for torsion. The maximum tensile strains of the polymer and composites are in the range of 5–93%, and their maximum torsional strains vary from 0.26 to 0.90. The composites exhibited very slow degradation rates in aqueous solution, from approximately 50% mass remaining for the pure polymer to 75% mass remaining for composites with high nHA contents, after a period of 8 weeks. Increase in ceramic content increased mechanical properties, but decreased maximum strain, degradation rate, and swelling of the composites. Human bone marrow stem cells and human endothelial cells adhered and proliferated on 4PLAUMA films and degradation products of the composites showed little-to-no toxicity. These results demonstrate that novel 4-arm poly(lactic acid urethane)-maleate (4PLAUMA) elastomer and its nHA composites may have potential applications in regenerative medicine.     

Front face fluorescence spectroscopy as a tool for the assessment of egg freshness during storage at a temperature of 12.2 degrees C and 87% relative humidity

The objective of this study was to investigate intrinsic fluorophores of thick albumen and egg yolk in order to assess egg freshness during storage at a temperature of 12.2 °C and 87% relative humidity (RH). A total of 126 intact brown-shelled eggs of the same flock (29 weeks of age) were stored for 1, 6, 8, 12, 15, 20, 22, 26, 29, 33, 40, 47 and 55 days. The emission fluorescence spectra of aromatic amino acids and nucleic acids (AAA + NA) (excitation: 250 nm; emission: 280-450 nm), fluorescent Maillard reaction products (FMRP) (excitation: 360 nm; emission: 380-580 nm) and the excitation spectra of vitamin A (emission: 410 nm; excitation: 270-350 nm) were scanned on thick albumen and egg yolk. Among the intrinsic fluorophores, only the principal component analysis (PCA) applied on the vitamin A fluorescence spectra allowed a good identification of eggs as a function of their storage time. Factorial discriminant analysis (FDA) was then applied on the first five principal components (PCs) of the PCA applied on each spectral data set. Regarding AAA + NA recorded on thick albumen, correct classification of 69.4% and 63.9% was observed for the calibration and validation sets, respectively. Quite similar results were obtained on AAA + NA scanned on egg yolks. The best results were obtained with vitamin A fluorescence spectra since 97.7% and 85.7% of the calibration and validation sets was obtained, respectively. These results showed that vitamin A fluorescence spectra provide useful fingerprints, mainly allow the identification of eggs during storage and could be considered as a powerful intrinsic probe for the evaluation of egg freshness.     

Rapid monitoring of grapevine reserves using ATR–FT-IR and chemometrics

Predictions of grapevine yield and the management of sugar accumulation and secondary metabolite production during berry ripening may be improved by monitoring nitrogen and starch reserves in the perennial parts of the vine. The standard method for determining nitrogen concentration in plant tissue is by combustion analysis, while enzymatic hydrolysis followed by glucose quantification is commonly used for starch. Attenuated total reflectance Fourier transform infrared spectroscopy (ATR–FT-IR) combined with chemometric modelling offers a rapid means for the determination of a range of analytes in powdered or ground samples. ATR–FT-IR offers significant advantages over combustion or enzymatic analysis of samples due to the simplicity of instrument operation, reproducibility and speed of data collection. In the present investigation, 1880 root and wood samples were collected from Shiraz, Semillon and Riesling vineyards in Australia and Germany. Nitrogen and starch concentrations were determined using standard analytical methods, and ATR–FT-IR spectra collected for each sample using a Bruker Alpha instrument. Samples were randomly assigned to either calibration or test data sets representing two thirds and one third of the samples respectively. Signal preprocessing included extended multiplicative scatter correction for water and carbon dioxide vapour, standard normal variate scaling with second derivative and variable selection prior to regression. Excellent predictive models for percent dry weight (DW) of nitrogen (range: 0.10–2.65% DW, median: 0.45% DW) and starch (range: 0.25–42.82% DW, median: 7.77% DW) using partial least squares (PLS) or support vector machine (SVM) analysis for linear and nonlinear regression respectively, were constructed and cross validated with low root mean square errors of prediction (RMSEP). Calibrations employing SVM-regression provided the optimum predictive models for nitrogen (R² = 0.98 and RMSEP = 0.07% DW) compared to PLS regression (R² = 0.97 and RMSEP = 0.08% DW). The best predictive models for starch was obtained using PLS regression (R² = 0.95 and RSMEP = 1.43% DW) compared to SVR (R² = 0.95; RMSEP = 1.56% DW). The RMSEP for both nitrogen and starch is below the reported seasonal flux for these analytes in Vitis vinifera. Nitrogen and starch concentrations in grapevine tissues can thus be accurately determined using ATR–FT-IR, providing a rapid method for monitoring vine reserve status under commercial grape production.     

Phase equilibria in ternary aqueous mixtures of 1,3-butanediol with 2-ethyl-1-hexanol at T = (298.2, 303.2 and 308.2) K

Experimental tie-line data for ternary system of (water + 1,3-butanediol (1,3-BD) + 2-ethyl-1-hexanol (2EH)) were determined at T = (298.2, 303.2 and 308.2) K under atmospheric conditions. This ternary system exhibits type-1 behavior of LLE. The experimental ternary LLE data were correlated using the NRTL model, and the binary interaction parameters were obtained. The average root-mean-square deviation between the observed and calculated mole fractions was 1.38%. Distribution coefficient and separation factor were measured to evaluate the extracting capability of the solvent. The separation factor values for the solvent used in this work were then compared with literature values obtained in our previous works for other butanediols.     

Determination of 15 isomers of chlorobenzoic acid in soil samples using accelerated sample extraction followed by liquid chromatography

A study was conducted to elaborate a fast, simple and efficient method for determination of 15 isomers chlorobenzoic acids (CBAs) in soil using HPLC-UV. Artificially contaminated soil samples were extracted using accelerated solvent extraction (ASE) with 1% acetic acid in a mixture of hexane and acetone (1:1, V/V) under a pressure of 10.34 MPa and temperature of 150 °C. The recovery of the ASE method was above 82%. The extracts were concentrated; dimethyl sulfoxide was used to prevent CBA volatilization and the final analysis was performed with a C18 XBridge HPLC column employing a mobile phase consisting of acetonitrile and 0.1% trifluoracetic acid in water. A HPLC procedure with gradient elution and UV detection was developed and validated. The method exhibited a linear range for 2-CBA; 2,6-CBA; 3-CBA; 4-CBA; 2,3-CBA; 2,3,6-CBA; 2,5-CBA; and 2,4-CBA from 5 to 120 μg/mL with a limit of quantification (LOQ) of 5 μg/mL, RSD from 2.42 to 9.42% and accuracy from 82 ± 2 to 103 ± 3%. The linear range of determination of 2,4,6-CBA, 3,4-CBA, 2,3,5,6-CBA, 3,5-CBA, 2,3,5-CBA, 2,3,4,5,6-CBA and 2,3,4,5-CBA was 10-120 μg/mL with LOQ 10 μg/mL, RSD from 0.74 to 5.84% and accuracy from 94 ± 1 to 114 ± 1%. The optimized analytical procedure was finally applied on two historically PCB contaminated soils and 9 CBAs were quantified in the samples.     

Interference-free determination of fluoroquinolone antibiotics in plasma by using excitation-emission matrix fluorescence coupled with second-order calibration algorithms

Fluoroquinolones or so-called second-generation quinolones, in particular, ofloxacin (OFL), norfloxacin (NOR), and enoxacin (ENO), with therapeutic advantages possess strongly overlapped fluorescence spectra. In this paper, two strategies were proposed for simultaneous direct determination of OFL, NOR and ENO in plasma by combining fluorescence excitation-emission matrix (EEM) with second-order calibration based on the alternating trilinear decomposition algorithm (ATLD) and parallel factor analysis (PARAFAC). The results showed that both algorithms could solve the problem of serious fluorescence spectral overlapping of the sought-for analytes even in the presence of uncalibrated interferents. However, ATLD has advantages of being insensitive to overestimated component number and fast convergence. The results by using ATLD with an estimated component number of five were reasonably acceptable for clinical analysis. The average recoveries of OFL, NOR and ENO in synthetic samples were 99.7 ± 2.4, 101.5 ± 2.4 and 97.3 ± 3.8%, respectively; the average recoveries of OFL, NOR and ENO in complex plasma were 94.3 ± 2.6, 85.6 ± 3.3 and 103.3 ± 3.0%, respectively.     

Measurement of thermal conductivity by differential scanning calorimetry

A technique of measurement of thermal conductivity of solid materials by differential scanning calorimetry is presented. It concerns small samples having a diameter less than 8.0 mm, a height less than 2.0 mm and a low thermal conductivity. This method requires many samples with different heights which are heated in such a way that a calibration substance put on their top undergoes a first-order phase transition. The analysis of heat transfer of a such experiment predicts that the slope of the differential power during the transition is proportional to the factor 2 and inversely proportional to the sum of the thermal resistances. A measurement of the thermal conductivity of samples made of polytetrafluoroethylene powder, compressed at the density of 2.10±0.03 g cm⁻³, has been performed; the value obtained is 0.33±0.02 W m⁻¹ K⁻¹. Measurements of thermal conductivity of small metal hydride pellets are also presented. The precision of the measurements are on average 10%.     

Liquid–liquid equilibria for mixtures containing water,methanol, fatty acid methyl esters,and glycerol

The liquid–liquid equilibrium (LLE) data, including tie-lines and phase boundaries, were measured for the ternary systems of water + methanol + methyl oleate, water + methanol + methyl linoleate, glycerol + methanol + methyl oleate, and glycerol + methanol + methyl linoleate at temperatures from 298.2 K to 318.2 K under atmospheric pressure. All the LLE data follow the Othmer-Tobias equation. Each ternary system behaves type-I LLE. The areas of two-liquid coexistence region decrease with increasing temperature. The experimental data were applied to test the validity of the UNIFAC model and its modified versions, including UNIFAC-LLE and UNIFAC-Dortmund. The LLE data were also correlated with the NRTL and the UNIQUAC models. The UNIQUAC model yielded better results.     

Liquid-liquid equilibria of water + 3-hydroxy-2-butanone + ethyl ethanoate

(Liquid-liquid) equilibrium (LLE) data of the solubility curves and tie-line compositions have been determined for mixtures of (water + 3-hydroxy-2-butanone + ethyl ethanoate) at 298.15 K, 308.15 K and 318.15 K and 101.3 kPa. Distribution coefficients and separation factors have been evaluated for the immiscibility region. The reliability of the experimental tie-lines has been confirmed by using Othmer-Tobias correlation. The LLE data of the ternary systems have been predicted by UNIFAC method.     

Discrimination of transgenic tomatoes based on visible/near-infrared spectra

VIS-NIR spectroscopy combined with multivariate analysis after the appropriate spectral data pre-treatment has been proved to be a very powerful tool for judgment of the relative pattern of the objects that have very similar properties. In this study, seventy transgenic tomatoes with antisense LeETR2 and 94 of their parents, non-transgenic ones were measured in VIS-NIR diffuse reflectance mode. Principal component analysis (PCA), discriminant analysis (DA) and partial least-squares discriminant analysis (PLSDA) were applied to classify tomatoes with different genes into two groups. Calibrations were developed using PLS regression with the leave-one-out cross-validation technique. The results show that differences between transgenic and non-transgenic tomatoes do exist and excellent classification can be obtained after optimizing spectral pre-treatment. The correct classifications for transgenic and non-transgenic tomatoes were both 100% using PLSDA after derivative spectral pre-treatment. The raw spectra with PLSDA model after the second derivative pre-treatment had the best satisfactory calibration and prediction abilities, with r_c = 0.97964, root mean square error of calibration (RMSEC) = 0.099, r_cv = 0.97963, root mean square error of cross-validation (RMSECV) = 0.0993 and a factor. The results in the present study show VIS-NIR spectroscopy together with chemometrics techniques could be used to differentiate transgenic tomato, which offers the benefit of avoiding time-consuming, costly and laborious chemical and sensory analysis.