Partial Least Squares Factors Research Articles

Model Optimization and Wavenumbers Selection for FTIR/ATR Spectroscopy Analysis of Glucose Aqueous Solution

The rapid quantification method of glucose aqueous solution was established by using the Fourier transform infrared spectroscopy (FTIR) and attenuated total reflection (ATR). Model optimization and wavenumbers selection was investigated based on Savitzky-Golay (SG) smoothing. For the whole spectral collecting region 4500-600 cm-1 and the fingerprint region 1600-900 cm-1, Partial least squares (PLS) models without and with SG smoothing were established respectively. The optimal model was on the fingerprint region with SG smoothing of 1st order derivative, 2nd degree polynomial and 73 smoothing points, PLS factor, RMSEP, RP, were 4, 0.331 mmol/L, 0.999 respectively. Based on 19 absorption peaks of the subtracted spectra of glucose aqueous solution for de-ionized water, all possible wavenumber combinations were used to establish discrete multiple linear regression (MLR) models respectively, the optimal wavenumbers combination was 3084, 1034, 991 (cm-1), RMSEP and RP were 0.459 mmol/L, 0.997 respectively. It could provide valuable references for designing spectrophotometer system in special spectrometer and further for glucose analysis of the complex system. To get the stable prediction results, all models and results here were obtained based on the average data on 50 different divisions of calibration set and prediction set.

A novel approach for development of a multivariate calibration model using a Doehlert experimental design: Application for prediction of key gasoline properties by Near-infrared Spectroscopy

Alternative methods for quality control in the petroleum industry have been obtained using Near-infrared Spectroscopy (NIRS) combined with multivariate techniques such as PLS (Partial Least-Square). The process of development and refinement of PLS models usually follows a nonsystematic and univariate procedure. The Standard Error of Cross Validation (SECV), the Standard Error of Prediction (SEP) and the determination coefficient (r 2 regr.) are usually the only guides used in pursuit of the best model. In the present work, a novel approach was proposed using a Doehlert experimental design with three input variables (wavenumber range, preprocessing technique and regression/validation technique) varied at 5, 7 and 3 levels, respectively. Besides SECV, SEP and r 2 regr., some additional response variables, such as the slope, r 2 and p value from the external validation, as well as the number of PLS factors, were simultaneously assessed to find the optimum conditions for PLS modeling. The optimum setting for each input variable was simultaneously defined through a multivariate approach using a desirability function. With the proposed approach, the main and interaction effects could also be investigated. The methodology was successfully applied to obtain PLS models to monitor the gasoline quality through the process of product loading in trucks. To prevent product contamination or adulteration, fast prediction of key properties was obtained from FT-NIR spectra within the 7300–3900 cm − 1 region with SECV in the range 0.04–0.63% w/w for composition (Aromatics, Saturates, Olefins and Benzene) and 0.0008 for Relative Density 20/4 °C. Each optimized PLS model was obtained with less than 40 modeling runs, demonstrating the efficiency of the proposed approach.

Waveband Selection of NIR Spectroscopy Analysis for Glucose Aqueous Solution Based on Savitzky-Golay Smoothing

The high accurate partial least squares (PLS) models of near-infrared (NIR) spectroscopy analysis for glucose aqueous solution were established by large-scale simultaneous optimization of Savitzky-Golay (SG) smoothing mode and PLS factor. A total of 6 wavebands were used for modelling respectively. The model prediction effect with SG smoothing was obviously better than ones without SG smoothing for each waveband, it indicated that the large-scale simultaneous optimization of SG smoothing mode and PLS factor could obviously improve model prediction effect. The optimal waveband was the combination of 400-1880 and 2082-2350 nm, the optimal SG smoothing mode was 1st order derivative smoothing, 2nd degree polynomial, 63 smoothing points, and the corresponding PLS factor, RMSEP and RP were 14, 0.289 mmol/L and 0.9996 respectively. The combination of 1100-1880 and 2082-2350 nm was also a waveband with high signal to noise ratio for glucose, could replace the whole spectral collecting region and get very good prediction effect, which could provide valuable references for designing spectrophotometer system in special NIR spectrometer. To get the stable prediction results, all models and results were obtained based on the average data on 30 different divisions of calibration set and prediction set.

Use of Airborne Hyperspectral Imagery to Map Soil Properties in Tilled Agricultural Fields

Soil hyperspectral reflectance imagery was obtained for six tilled (soil) agricultural fields using an airborne imaging spectrometer (400–2450 nm,∼10 nm resolution, 2.5 m spatial resolution). Surface soil samples (n=315) were analyzed for carbon content, particle size distribution, and 15 agronomically important elements (Mehlich-III extraction). When partial least squares (PLS) regression of imagery-derived reflectance spectra was used to predict analyte concentrations, 13 of the 19 analytes were predicted withR2>0.50, including carbon (0.65), aluminum (0.76), iron (0.75), and silt content (0.79). Comparison of 15 spectral math preprocessing treatments showed that a simple first derivative worked well for nearly all analytes. The resulting PLS factors were exported as a vector of coefficients and used to calculate predicted maps of soil properties for each field. Image smoothing with a3×3low-pass filter prior to spectral data extraction improved prediction accuracy. The resulting raster maps showed variation associated with topographic factors, indicating the effect of soil redistribution and moisture regime on in-field spatial variability. High-resolution maps of soil analyte concentrations can be used to improve precision environmental management of farmlands.

ORIGINAL ARTICLE: Assessment of mildew levels in wheat samples based on spectral characteristics of bulk grains

Background Mildew damage is a quality defect that has an adverse effect on the quality of flour. Aims The objective of this study was to develop a rapid and consistent imaging method to quantify the extent of mildew damage in wheat samples. Materials and methods A hyperspectral imaging system with a wavelength range of 400–1000 nm was used to detect and quantify mildew in 65 Canada Eastern Soft Red Winter (CESRW) wheat samples. Partial least square (PLS) regression calibrations were developed to predict mildew levels based on the spectral characteristics of the bulk samples. Results and discussion Predictions from a model with 4 PLS factors based on image standard deviation spectra matched well with the visual assessment of the samples with an R2 approaching 0.87 and an RMSE of 0.76 on the validation set. Accuracy of the PLS classification for 9 mildew levels was 90.6% (±1 level) and 84.4% for 3 inspector grades. Conclusion This study confirms that potential use of hyperspectral imaging for mildew detection in commercial operations is possible. Shahin MA, Hatcher, DW, Symons JS (2010). Assessment of mildew levels in wheat samples based on spectral characteristics of bulk grains. Quality Assurance and Safety of Crops & Foods, 2, 133–140.

Acetaminophen determination in low-dose pharmaceutical syrup by NIR spectroscopy

The aim of the present study was first to develop a robust near infrared (NIR) calibration model able to determine the acetaminophen content of a low-dose syrup formulation (2%, w/v). Therefore, variability sources such as production campaigns, batches, API concentration, syrup basis, operators and sample temperatures were introduced in the calibration set. A prediction model was then built using partial least square (PLS) regression. First derivative followed by standard normal variate (SNV) were chosen as signal pre-processing. Based on the random subsets cross-validation, 4 PLS factors were selected for the prediction model. The method was then validated for an API concentration ranging from 16 to 24 mg/mL (1.6–2.4%, w/v) using an external validation set. The 0.26 mg/mL RMSEP suggested the global accuracy of the model. The accuracy profile obtained from the validation results, based on tolerance intervals, confirmed the adequate accuracy of the results generated by the method all over the investigated API concentration range. Finally, the NIR model was used to monitor in real time the API concentration while mixing syrups containing various amounts of API, a good agreement was found between the NIR method and the theoretical concentrations.

Learning "graph-mer" motifs that predict gene expression trajectories in development.

A key problem in understanding transcriptional regulatory networks is deciphering what cis regulatory logic is encoded in gene promoter sequences and how this sequence information maps to expression. A typical computational approach to this problem involves clustering genes by their expression profiles and then searching for overrepresented motifs in the promoter sequences of genes in a cluster. However, genes with similar expression profiles may be controlled by distinct regulatory programs. Moreover, if many gene expression profiles in a data set are highly correlated, as in the case of whole organism developmental time series, it may be difficult to resolve fine-grained clusters in the first place. We present a predictive framework for modeling the natural flow of information, from promoter sequence to expression, to learn cis regulatory motifs and characterize gene expression patterns in developmental time courses. We introduce a cluster-free algorithm based on a graph-regularized version of partial least squares (PLS) regression to learn sequence patterns—represented by graphs of k-mers, or “graph-mers”—that predict gene expression trajectories. Applying the approach to wildtype germline development in Caenorhabditis elegans, we found that the first and second latent PLS factors mapped to expression profiles for oocyte and sperm genes, respectively. We extracted both known and novel motifs from the graph-mers associated to these germline-specific patterns, including novel CG-rich motifs specific to oocyte genes. We found evidence supporting the functional relevance of these putative regulatory elements through analysis of positional bias, motif conservation and in situ gene expression. This study demonstrates that our regression model can learn biologically meaningful latent structure and identify potentially functional motifs from subtle developmental time course expression data.

Application of near infrared spectroscopy for rapid analysis of intermediates of Tanreqing injection

A method for rapid quantitative analysis of four kinds of Tanreqing injection intermediates was developed based on Fourier transform near infrared (FT-NIR) spectroscopy and partial least squares (PLS) algorithm. The NIR spectra of 120 samples were collected in transflective mode. The concentrations of chlorogenic acid, caffeic acid, luteoloside, baicalin, ursodesoxycholic acid (UDCA), and chenodeoxycholic acid (CDCA) were determined with the HPLC–DAD/ELSD as reference method. In the PLS calibration, the NIR spectra were pretreated with different methods and the number of PLS factors used in the model calibration was optimized by leave-one-out cross-validation. The performance of the final PLS models was evaluated according to the root mean square error of calibration (RMSEC), root mean square error of cross-validation (RMSECV), root mean square error of prediction (RMSEP), BIAS, standard error of prediction (SEP), and correlation coefficients ( R). The R values in the prediction sets were all higher than 0.93, and the SEPs for the 6 compounds are 1.18, 6.02, 2.71, 155, 126, 30.0 mg/l, respectively. The established models were used for the liquid preparation process analysis of Tanreqing injection in three batches, and a model updating method was proposed for the long-term usage of the established models. This work demonstrated that NIR spectroscopy is more rapid and convenient than the conventional methods to analyze the intermediates of Tanreqing injection, and the presented method is helpful to the implementation of process analytical technology (PAT) in pharmaceutical industry of Chinese Medicines Injections.

Multivariate calibration of spectrophotometric data using a partial least squares with data fusion

A novel method named DF-PLS based on partial least squares (PLS) regression combined with data fusion (DF) was applied to enhance the ability of extracting characteristic information and the quality of regression for the simultaneous spectrophotometric determination of Cu(II), Ni(II) and Cr(III). Data fusion is a technique that seamlessly integrates information from disparate sources to produce a single model or decision. Wavelet representations of signals provide a local time-frequency description and are multiscale in nature, thus in the wavelet domain, the quality of noise removal is implemented by a scale-dependent threshold method. Information from different wavelet scales is just like different sources of information. Integrating the information from different wavelet scales to obtain a PLS model belongs to the technique of data fusion. PLS was applied for multivariate calibration and noise reduction by eliminating the less important latent variables. In this case, by optimization, wavelet functions, decomposition level and thresholding methods and the number of PLS factors for DF-PLS were selected as Daubechies 4, 7, HYBRID thresholding and 3, respectively. The relative standard errors of prediction (RSEP) for all compounds with DF-PLS and PLS were 3.13% and 10.3%, respectively. Experimental results showed the DF-PLS method to be successful for simultaneous multicomponent determination even when severe overlap of spectra was present and proved it to be better than PLS. The DF-PLS method is a hybrid technique that combines the best attributes of DF and PLS, which makes it a promising and attractive method.

Rapid quantification of phenolic acids in Radix Salvia Miltrorrhiza extract solutions by FT-NIR spectroscopy in transflective mode

A rapid method for simultaneous determination of main phenolic acids in Radix Salvia Miltrorrhiza extract solutions was developed using Fourier transform near infrared spectroscopy in transflective mode and multivariate calibration and HPLC-UV as the reference method. Partial least squares (PLS) algorithm was conducted on the calibration of regression models. The multiplicative scatter correction, Norris derivative and second derivative were adopted for the spectral pre-processing, and the number of PLS factors were optimized by leave-one-out cross-validation. The performance of the final model was evaluated according to root mean square error of prediction (RMSEP) and correlation coefficient ( R). The R values achieved in the prediction set were above 0.93. The developed models were used for analysis of unknown samples and routine monitoring with satisfactory results. This work demonstrated that NIR spectroscopy combined with PLS algorithm could be used for the rapid determination of the main phenolic acids of Salvia Miltrorrhiza extract solutions.

Development of near infrared reflectance spectroscopy (NIRS) calibration model for estimation of oil content in a worldwide safflower germplasm collection

The development of NIRS calibration model as a rapid, precise, robust, and cost-effective method to estimate oil content in ground seeds of worldwide safflower germplasm collection grown under different agro-climatic conditions was the key objective of this research project. The oil content was measured by accelerated solvent extraction method in a total of 328 samples collected across 2004 (165 samples) and 2005 (163) growing seasons and used as reference values. Two thirds of the measured samples were used for building the calibrations and one third for the validations. Combined and annual calibration and validation models were carried out by NIRCal 4.21 using the partial least squares (PLS) regression. Different data pretreatments such as full multiplicative scatter correction (MSC), first derivative or smoothing way of Savitzky-Golay with a gap of 9 data points were used to improve the calibration models. The optimum PLS factors for developing the best calibration were 12, 10, and 14 for combined model, annual model of 2004 and of 2005, respectively. In combined and annual models, the statistical parameters in calibration model were consistent with the respective parameters in validation model. Coefficient of variation (15.5 to 25.1) demonstrated high variability in calibration and validation models. The standard error of estimation (SEE) and standard error of prediction (SEP) for combined model were 1.40 and 1.43, respectively. Although the quality value (Q-value) of calibration was slightly higher in annual models (0.66 for both), the combined calibration model (0.64) precisely predicted oil content as indicated by higher coefficient of determination (0.90) and RPD (3.2%) compared to annual calibration. The accuracy and precision of the combined calibration model were sufficient to use NIRS as a tool for screening of oil content in a diverse safflower germplasm in the range obtained.

Determination of potency of heparin active pharmaceutical ingredient by near infrared reflectance spectroscopy

Potency is an important parameter for evaluation of quality of heparin active pharmaceutical ingredient (API). In this paper the feasibility to determine potency of heparin API with near infrared reflectance spectroscopy (NIRS) coupled with partial least squares (PLS) algorithm is attempted. PLS factors, correlation coefficient of calibration set ( R c), the root mean square of cross-validation (RMSECV), correlation coefficient of prediction set ( R p) and the root mean square of prediction (RMSEP) were used to evaluate the performance of the models. The optimal calibration model was obtained with R p = 0.9721 and RMSEP = 0.55 in the 1700–1898 nm spectral region when using SG-1st derivative spectral transform method and division of calibration/prediction samples was 1/1. Three other additional samples demonstrated good prediction capability of the final model and three validation samples gave good repeatability result. NIRS has the potential to be a final lot release test to be performed in a QC laboratory.

Active content determination of non-coated pharmaceutical pellets by near infrared spectroscopy: Method development, validation and reliability evaluation

A robust near infrared (NIR) method able to quantify the active content of pilot non-coated pharmaceutical pellets was developed. A protocol of calibration was followed, involving 2 operators, independent pilot batches of non-coated pharmaceutical pellets and two different NIR acquisition temperatures. Prediction models based on Partial Least Squares (PLS) regression were then carried out. Afterwards, the NIR method was fully validated for an active content ranging from 80 to 120% of the usual active content using new independent pilot batches to evaluate the adequacy of the method to its final purpose. Conventional criteria such as the R(2), the Root Mean Square Error of Calibration (RMSEC), the Root Mean Square Error of Prediction (RMSEP) and the number of PLS factors enabled the selection of models with good predictive potential. However, such criteria sometimes fail to choose the most fitted for purpose model. Therefore, a novel approach based on accuracy profiles of the validation results was used, providing a visual representation of the actual and future performances of the models. Following this approach, the prediction model using signal pre-treatment Multiplicative Scatter Correction (MSC) was chosen as it showed the best ability to quantify accurately the active content over the 80-120% active content range. The reliability of the NIR method was tested with new pilot batches of non-coated pharmaceutical pellets containing 90 and 110% of the usual active content, with blends of validation batches and industrial batches. All those batches were also analyzed by the HPLC reference method and relative errors were calculated: the results showed low relative errors in full accordance with the results obtained during the validation of the method, indicating the reliability of the NIR method and its interchangeability with the HPLC reference method.

Pharmacophore generation and atom-based 3D-QSAR of novel 2-(4-methylsulfonylphenyl)pyrimidines as COX-2 inhibitors

Cyclooxygenase-2 (COX-2) inhibitors are widely used for the treatment of pain and inflammatory disorders such as rheumatoid arthritis and osteoarthritis. A series of novel 2-(4-methylsulfonylphenyl)pyrimidine derivatives has been reported as COX-2 inhibitors. In order to understand the structural requirement of these COX-2 inhibitors, a ligand-based pharmacophore and atom-based 3D-QSAR model have been developed. A five-point pharmacophore with four hydrogen bond acceptors (A) and one hydrogen bond donor (D) was obtained. The pharmacophore hypothesis yielded a 3D-QSAR model with good partial least-square (PLS) statistics results. The training set correlation is characterized by PLS factors (r (2) = 0.642, SD = 0.65, F = 82.7, P = 7.617 e - 12). The test set correlation is characterized by PLS factors (Q (2) (ext) = 0.841, RMSE = 0.24,Pearson-R = 0.91). A docking study revealed the binding orientations of these inhibitors at active site amino acid residues (Arg513, Val523, Phe518, Ser530, Tyr355, His90) of COX-2 enzyme. The results of ligand-based pharmacophore hypothesis and atom-based 3D-QSAR give detailed structural insights as well as highlights important binding features of novel 2-(4-methylsulfonylphenyl)pyrimidine derivatives as COX-2 inhibitors which can provide guidance for the rational design of novel potent COX-2 inhibitors.

Multivariate calibration of on-line enrichment near-infrared (NIR) spectra and determination of trace lead in water

An on-line enrichment near-infrared spectroscopy (NIRS) technique was developed in order to improve the sensitivity for the determination of trace heavy metal lead. Trace Pb 2+ was enriched by chelating resin with on-line enrichment equipment and then the enriched lead on the resin was directly determined by near-infrared diffuse reflectance spectroscopy without elution procedure for simplicity of measurements. Partial least squares (PLS) regression was used to build models between the corrected spectra and concentrations of Pb 2+, and segmental cross-validation was used to search for a reasonable number of PLS factors. The results showed that NIR spectra and concentrations of Pb 2+ had a good linear relationship with the maximum of correlation coefficient of 0.9228 in the 1006–1383 nm region. The minimum RMSEP was 0.4777 mg L − 1 when the latent variable number was 5 by using the region of 1006–1383 nm. The limit of detection (LOD) of lead was 0.2384 mg L − 1 in this case. This study demonstrates that the prediction of trace heavy metal lead concentrations by NIRS is feasible with enrichment technique.

Combining orthogonal signal correction and wavelet packet transform with partial least squares to analyze overlapping spectra of three kinds of metal ions

A novel method named OSC-WPT-PLS approach based on partial least squares (PLS) regression with orthogonal signal correction (OSC) and wavelet packet transform (WPT) as pre-processed tools was proposed for the simultaneous spectrophotometric determination of Al(III), Mn(II) and Co(II). This method combines the ideas of OSC and WPT with PLS regression for enhancing the ability of extracting characteristic information and the quality of regression. OSC is used to remove information in the response matrix D by subtracting the structured noise that is orthogonal to the concentration matrix C. Wavelet packet transform was applied to perform data compression, to extract relevant information, and to eliminate noise and collinearity. PLS was applied for multivariate calibration and noise reduction by eliminating the less important latent variables. In this case, using trials, the kind of wavelet function, the decomposition level, the number of OSC components and the number of PLS factors for the OSC-WPT-PLS method were selected as Daubechies 4, 3, 2 and 3, respectively. A program (POSCWPTPLS) was designed to perform the simultaneous spectrophotometric determination of Al(III), Mn(II) and Co(II). The relative standard errors of prediction (RSEP) obtained for total elements using OSC-WPT-PLS, WPT-PLS and PLS were compared. Experimental results demonstrated that the OSC-WPT-PLS method had the best performance among the three methods and was successful even when there was severe overlap of spectra.

Moisture content determination of pharmaceutical pellets by near infrared spectroscopy: Method development and validation

The aim of the present study was to develop and validate a near infrared method able to accurately determine a moisture content of pharmaceutical pellets ranging from 1% to 8% in order to check their moisture content conformity. A calibration and validation set were designed for the conception and evaluation of the method adequacy. An experimental protocol was then followed, involving two operators, independent production campaign batches and different temperatures for data acquisition. On the basis of this protocol, prediction models based on partial least squares (PLS) regression were then carried out. Conventional criteria such as the R 2, the root mean square errors of calibration and prediction (RMSEC and RMSEP) as well as the number of PLS factors enabled the selection of three preliminary models. However, such criteria did not clearly demonstrate the model's ability to give accurate predictions over the whole analyzed water content range. Consequently, a novel approach based on accuracy profiles which allow the selection of the most fitted model for purpose was used. According to this novel approach, the model using multiplicative scatter correction (MSC) pre-treatment was obviously the most suitable. Indeed, the resulting accuracy profile clearly showed that this model was able to determine moisture content over the range of 1–8% with a very acceptable accuracy. The present study confirmed that NIR spectroscopy could be used in the PAT concept as a non-invasive, non-destructive and fast technique for moisture content determination in pharmaceutical pellets. In addition, facing the limit of the classical and commonly used criteria, the use of accuracy profiles proved to be useful as a powerful decision tool to demonstrate the suitability of the proposed analytical method.

Combining direct orthogonal signal correction and wavelet packet transform with partial least squares to analyze overlapping voltammograms of nitroaniline isomers

A novel method named DOSCWPTPLS approach based on partial least squares (PLS) regression with direct orthogonal signal correction (DOSC) and wavelet packet transform (WPT) as pre-processed tools was proposed for the simultaneous differential pulse voltammetric determination of o-nitroaniline, m-nitroaniline and p-nitroaniline with overlapping peaks. The method combines the ideas of DOSC and WPT with PLS regression for enhancing the ability in the extraction of characteristic information and the quality of regression. Data reduction was performed using DOSC, WPT and PLS algorithm. DOSC is used to remove information in the response matrix D by subtracting the structured noise that is orthogonal to the concentration matrix C. Wavelet packet representations of signals provide a local time–frequency description, thus in the wavelet packet domain, the quality of the noise removal, data compression, and relevant information extraction can be improved. PLS was applied for multivariate calibration and reduced noise by eliminating the less important latent variables. In this case, by optimization, the kind of wavelet function, the decomposition level, and the number of DOSC components, tolerance factor and the number of PLS factors for the DOSCWPTPLS method were selected as Coiflet 2, 3, 3, 0.001 and 3, respectively. A program (PDOSCWPTPLS) was designed to perform the simultaneous voltammetric determination of o-nitroaniline, m-nitroaniline and p-nitroaniline. The relative standard errors of prediction (RSEP) obtained for all components using DOSCWPTPLS, DOSCPLS, WPTPLS and PLS were compared. Experimental results demonstrated that the DOSCWPTPLS method had the best result among the four methods and was successful even when there was severe overlap of voltammograms.

Multimode Methods Applied on MIA Descriptors in QSAR

Since the introduction of physicochemical descriptors to derive useful QSAR (quantitative structure-activity relationship) models, some regression methods have been applied to linearly correlate dependent (bioactivities) and independent variables. Multiple linear regression (MLR) has been widely used when the number of samples (rows) exceed the amount of descriptors (columns), whilst partial least squares (PLS) is the most commonly applied regression method in 3D QSAR (e.g. CoMFA and related methods), where a large number of descriptors are generated. The recently implemented MIA-QSAR (Multivariate Image Analysis applied to QSAR) method is a especial (not only) case in which the descriptors (pixels) for each active compound result in a three-way array after grouping samples to give a data set. Such array may be properly treated by using N-way methods, such as multilinear PLS (N-PLS) and parallel factor analysis (PARAFAC). However, these methods have not been appropriately explored in QSAR studies, despite their supposed advantages over well established methods. Thus, this review formally details the MIA-QSAR approach prior to presenting two promising multimode methods to be applied on MIA descriptors, namely N-PLS and PARAFAC. Also, the suitability of such methods is discussed in terms of application to a case study (a series of anti-HIV compounds) and comparison to traditional (bilinear) PLS and docking studies.

The quantified analysis of fresh mutton tenderness using PLS methods and Fourier transform near-infrared spectroscopy

Ninety eight representative fresh mutton samples from Neimeng, Ningxia, Gansu, Xinjiang province were selected for this study, the nondestructive measurement of the fresh mutton tenderness by Fourier transform near infrared (FT-NIR) spectroscopy was discussed. Partial least squares(PLS) algorithm was used to build the model between the shear force value of the fresh mutton tenderness measured by the texture machine and the FT-NIR spectra. The influence of different processing method of spectra, factors and wave regions on the determination coefficients (r2), root mean square error of cross validation (RMSECV) and root mean square error of prediction (RMSEP) was studied. The result showed that the shear force value of ninety eight representative fresh mutton samples was 1.673-6.631 kg, and the shear force value above 75% samples was 2-5 kg, almost covering the fresh mutton tenderness of our country's sheep, the r2 of the calibration could reach 86.2% and the RMSECV was up to 0. 445 in the wave number range 11 995-5 446 cm(-1) and 4 601-4 246 cm(-1) with vector normalization when the PLS factors was ten. The correlation coefficient(R), RMSEP and average bias between value measured by the texture machine and predicted value of model based on validation samples were 0.87, 0.524 and 0.385 respectively. The result indicates that FT-NIR spectroscopy is capable of predicting tenderness value of fresh mutton.