Cross-validation Coefficient Of Determination Research Articles

Non-invasive lactate- and pH-monitoring in porcine meat using Raman spectroscopy and chemometrics

The feasibility of using chemometrics and Raman spectroscopy as a fast and non-invasive method to monitor the early postmortem lactate accumulation and pH decline in pork meat has been investigated. For this application, an on-line monitoring methodology has not yet been established. Based on raw Raman spectra of porcine semimembranosus muscles, a range of spectral pre-processing and multivariate calibration techniques were investigated to develop and test on-line prediction models for the meat quality parameters. The influence of the pre-processing methods on the prediction speed, robustness and accuracy performance of the employed linear and non-linear algorithms was compared. Identification of the most effective chemometric evaluation procedure was performed using least square linear regression together with locally weighted regression and metaheuristic data optimization methods such as the genetic algorithm and the ant colony optimization. The herein presented analysis suggests that the locally weighted regression applied to the standard normal variate (SNV) normalized Raman spectra provides the most accurate and robust models with a cross-validated coefficient of determination (r2cv) of 0.97 for pH and lactate, a cross-validated root mean square error (RMSECV) of 4.5mmol/kg for the lactate prediction and 0.06pH-units for the pH prediction. These results demonstrate the great potential of combining chemometrics and Raman spectroscopy for on-line meat quality control applications.

Predicción de la calidad de lana mediante Espectroscopia de Reflectancia en el Infrarrojo Cercano (NIRS)

This study was performed to evaluate the potential of Near Infrared Spectroscopy (NIRS) to predict some economically relevant variables of sheep wool, such as average fiber diameter (DMF), traction resistance (RT), scouring percentage (RL), comfort factor (FC) and staple length (LM). Corriedale and Multi Purpose Merino (MPM) wool samples (22 of each genotype) were scanned with a NIRSystems 6500 monochromator. To obtain the reference data, samples were analyzed with an OFDA 2000 instrument (Consorcio Ovino S.A.) and using the Sirolan Laser scan technique (New Zealand Wool Testing Authority, NZWTA Laboratory). Regression equations (WINISI II software) were obtained for the spectral data and the different variables of wool quality, by using modified partial least squares (MPLS) applying different mathematical treatments to the spectral values. The best equations were selected based on the determination coefficient of cross validation (R2CV), the standard error of cross validation (SECV) and the relation RPD, i.e., between standard deviation of reference data and the SECV. Three out of the five wool variables (DMF, RL and FC) could be reliably predicted by the NIR spectra, with calibrations that achieved a RPD>3 (DMF and FC) or very close to 3 (RL). No reliable equations could be obtained for RT and LM. It is concluded that NIRS has the capability to predict some wool variables of economic relevance, such as fiber diameter, comfort factor and scouring percentage.

Quantitative structure-activity relationship model for the fetal-maternal blood concentration ratio of chemicals in humans.

A quantitative structure-activity relationship (QSAR) model of the fetal-maternal blood concentration ratio (F/M ratio) of chemicals was developed to predict the placental transfer in humans. Data on F/M ratio of 55 compounds found in the literature were separated into training (75%, 41 compounds) and testing sets (25%, 14 compounds). The training sets were then subjected to multiple linear regression analysis using the descriptors of molecular weight (MW), topological polar surface area (TopoPSA), and maximum E-state of hydrogen atom (Hmax). Multiple linear regression analysis and a cross-validation showed a relatively high adjusted coefficient of determination (Ra(2)) (0.73) and cross-validated coefficient of determination (Q(2)) (0.71), after removing three outliers. In the external validation, R(2) for external validation (R(2)pred) was calculated to be 0.51. These results suggested that the QSAR model developed in this study can be considered reliable in terms of its robustness and predictive performance. Since it is difficult to examine the F/M ratio in humans experimentally, this QSAR model for prediction of the placental transfer of chemicals in humans could be useful in risk assessment of chemicals in humans.

QSAR models for thiophene and imidazopyridine derivatives inhibitors of the Polo-Like Kinase 1

The inhibitory activity of 103 thiophene and 33 imidazopyridine derivatives against Polo-Like Kinase 1 (PLK1) expressed as pIC50 (−logIC50) was predicted by QSAR modeling. Multivariate linear regression (MLR) was employed to model the relationship between 0D and 3D molecular descriptors and biological activities of molecules using the replacement method (MR) as variable selection tool.The 136 compounds were separated into several training and test sets. Two splitting approaches, distribution of biological data and structural diversity, and the statistical experimental design procedure D-optimal distance were applied to the dataset. The significance of the training set models was confirmed by statistically higher values of the internal leave one out cross-validated coefficient of determination (Q2) and external predictive coefficient of determination for the test set (Rtest2).The model developed from a training set, obtained with the D-optimal distance protocol and using 3D descriptor space along with activity values, separated chemical features that allowed to distinguish high and low pIC50 values reasonably well. Then, we verified that such model was sufficient to reliably and accurately predict the activity of external diverse structures.The model robustness was properly characterized by means of standard procedures and their applicability domain (AD) was analyzed by leverage method.

Potential of spectroscopic techniques and chemometric analysis for rapid measurement of docosahexaenoic acid and eicosapentaenoic acid in algal oil

Developing rapid methods for measuring long-chain ω-3 (n-3) poly-unsaturated fatty acid (LCPUFA) contents has been a crucial request from the algal oil industry. In this study, four spectroscopy techniques, namely visible and short-wave near infra-red (Vis–SNIR), long-wave near infra-red (LNIR), mid-infra-red (MIR) and nuclear magnetic resonance (NMR) spectroscopy, were exploited for determining the docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) contents in algal oil. The best prediction for both DHA and EPA were achieved by NMR spectroscopy, in which the determination coefficients of cross-validation (rCV2) values were 0.963 and 0.967 for two LCPUFAs. The performances of Vis–SNIR and LNIR spectroscopy were also accepted. The variable selection was proved as an efficient and necessary step for the spectral analysis in this study. The results were promising and implied that spectroscopy techniques have a great potential for assessment of DHA and EPA in algal oil.

基于最佳波段判别的湿地植物叶片全氮反演研究

PDF HTML阅读 XML下载 导出引用 引用提醒 基于最佳波段判别的湿地植物叶片全氮反演研究 DOI: 10.5846/stxb201311042664 作者: 作者单位: 首都师范大学资源环境与旅游学院,首都师范大学资源环境与旅游学院,首都师范大学资源环境与旅游学院,首都师范大学资源环境与旅游学院 作者简介: 通讯作者: 中图分类号: 基金项目: 国家自然科学基金（40901281）；北京市教育委员会科技计划面上项目（KM201310028012） Estimating Wetland plant leaf total nitrogen content based on optimal bands of reflectance from wetland vegetation Author: Affiliation: College of Resources Environment and Tourism,Capital Normal University,College of Resources Environment and Tourism,Capital Normal University,College of Resources Environment and Tourism,Capital Normal University,College of Resources Environment and Tourism,Capital Normal University Fund Project: 摘要 | 图/表 | 访问统计 | 参考文献 | 相似文献 | 引证文献 | 资源附件 | 文章评论 摘要:利用高光谱遥感技术定量估测湿地植被叶片全氮含量，对于监测和诊断湿地植被的生理状况及生长趋势具有重要意义。但叶片氮素遥感诊断研究多存在反演模型过拟合、入选波段与生化参量间因果关系不明确和入选变量间"多重共线性"等局限。以芦苇（Phragmites australis）和香蒲（Typha angustifolia）叶片全氮含量作为研究对象，通过谱带分区，分区最佳波段选取和偏最小二乘回归相结合的方法构建芦苇和香蒲叶片全氮含量反演模型，并利用交叉验证决定系数（Rcv2）和均方根误差（RMSEcv）对模型精度进行检验，尝试克服传统反演方法中的不足。结果表明，不同湿地植物类型相比，利用芦苇反射光谱建立的预测模型精度都高于香蒲。不同回归模型相比，一阶导数光谱-偏最小二乘回归模型（FDS-PLSR）精度远高于原始光谱-偏最小二乘回归模型（OS-PLSR）。芦苇最佳模型交叉验证决定系数（Rcv2）达到了0.84，方根误差（RMSEcv）为0.10，香蒲最佳模型交叉验证的决定系数（Rcv2）达到了0.66，方根误差（RMSEcv）为0.13，是构建湿地植物芦苇和香蒲光谱与叶片全氮含量关系的最佳模型。在不降低湿地植物叶片氮含量反演精度的基础上，有效地避免了传统地物高光谱模型反演中的局限性，是无损害遥感探测方面的有益尝试。 Abstract:Utilization of hyperspectral remote sensing technology to estimate wetland plant leaf nitrogen content quantitatively in large area is important to monitor and diagnosis of physiological condition and growth trend of wetland vegetation. However, there are many limitations such as over-fitting of inversion model, indeterminate causal relationship between the selected bands and biochemical parameters,"multicollinearity" of selected bands in leaf nitrogen diagnosis remote sensing research. The total nitrogen content of leaves of typical wetland plants, Phragmites australis and Typha angustifolia, was selected as our study objects. These plants grow in South Wetland purification system in the Olympic Park in Beijing, a typical wetland using reclaimed water. The leaf reflectance spectra of main wetland plants were acquired by means of an ASD FieldSpec 3 spectrometer (350-2500nm). Leaf total nitrogen content was determined by Kjeldahl nitrogen measurement method after acquiring the leaf reflectance spectra. The method,"Give a cross correlation analysis", was used to build the correlations between leaf nitrogen content and the original spectrum, first derivative spectrum. Then, the selected bands were divided into some areas according to QiSeGuang spectral range and the interval of selected bands. Those bands which are high frequency、frequency more、ranked high frequency are chosen as representative bands of different areas and considered to be the optimal bands used to build the regression model. Finally,partial least squares was used to build inversion model. The accuracy of this model was tested with cross-validated coefficient of determination (Rcv2) and cross、|validated root mean square error (RMSEcv).The results show that the first derivative transformation can effectively improve the sensitivity of the original spectrum leaf nitrogen content inversion, and fully reflect the sensitivity of near infrared wave band representing leaf total nitrogen content. The accuracy of regression model based on the first derivative and the partial least-squares was much higher than that of the original spectra. In the regression model of the reed, verification accuracy (Rcv2) reached 0.84, square root error (RMSEcv) was 0.11, in the regression model of the cattail, verification accuracy (Rcv2) reached 0.66, square root error (RMSEcv) was 0.13, which were the optimal models to estimate leaf total nitrogen content. The determination of parameters in "Give a cross correlation analysis" provided a scientific basis for building the model of eliminating "singularband" and reducing "multicollinearity" problem. Spectrum zone division provides a scientific basis for revealing the causal relationships between optimal band and biochemical parameters. And partial least squares regression method was used to avoid "multicollinearity" of selected bands. The result from this study can not only fill the gaps in the detection of leaf nitrogen using remote sensing, but also provide a strong scientific basis for the nitrogen content monitoring and management of urban wetlands using reclaimed water. At last, Partial least squares regression method was used to avoid "multicollinearity" of selected bands. 参考文献 相似文献 引证文献

ANN Model to Predict the Influence of Chemical and Biological Parameters on Iron and Manganese Accumulation

Water companies have adopted sophisticated risk-based management systems for managing water quality in water distribution networks (WDNs). Despite their efforts to comply with the standards for drinking water, they continue to receive customer complaints related to the water quality; discoloration is one such customer complaint. These complaints greatly undermine customers’ confidence in water companies. Discoloration is the result of release of accumulated material on pipe walls under stressed conditions. Therefore, understanding the causes of accumulation and the processes that influence accumulation of material is of paramount importance to water companies. In this paper, initially we identified various chemical and biological processes that highly influence the process of accumulation. Thereafter, using six years of water quality data, collected randomly, an artificial neural network (ANN) model was developed to predict Iron (Fe) and Manganese (Mn) accumulation potential. From the prediction profiler graph of the model, it was observed that increasing aluminium in the range 0 to 120μg/l resulted in an increase in Fe and Mn accumulation potential due to increased sorption capabilities. It was also observed that free chlorine residual FCR has a dual effect on Fe and Mn accumulation potential. A cross-validation coefficient of determination, R2 of 0.70 for the ANN model indicates that the model is likely to predict accumulation potential well on new datasets.

Mathematical relationships for metal build-up on urban road surfaces based on traffic and land use characteristics

The study investigated the influence of traffic and land use parameters on metal build-up on urban road surfaces. Mathematical relationships were developed to predict metals originating from fuel combustion and vehicle wear. The analysis undertaken found that nickel and chromium originate from exhaust emissions, lead, copper and zinc from vehicle wear, cadmium from both exhaust and wear and manganese from geogenic sources. Land use does not demonstrate a clear pattern in relation to the metal build-up process, though its inherent characteristics such as traffic activities exert influence. The equation derived for fuel related metal load has high cross-validated coefficient of determination (Q2) and low Standard Error of Cross-Validation (SECV) values which indicates that the model is reliable, while the equation derived for wear-related metal load has low Q2 and high SECV values suggesting its use only in preliminary investigations. Relative Prediction Error values for both equations are considered to be well within the error limits for a complex system such as an urban road surface. These equations will be beneficial for developing reliable stormwater treatment strategies in urban areas which specifically focus on mitigation of metal pollution.

Comparison of Infrared Spectroscopy and Nuclear Magnetic Resonance Techniques in Tandem with Multivariable Selection for Rapid Determination of ω-3 Polyunsaturated Fatty Acids in Fish Oil

This study investigated the potential of using four spectroscopic techniques including visible–short-wave near infrared, long-wave near infrared (LNIR), mid-infrared, and nuclear magnetic resonance (NMR) spectroscopy in tandem with multivariable selection and calibration for rapid determination of three important ω-3 polyunsaturated fatty acids (PUFA), namely eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), and docosapentaenoic acid (DPA) in fish oil. Quantitative models were established between the spectral data and reference PUFA contents of samples based on partial least squares regression (PLSR) algorithm. Successive projections algorithm (SPA) and uninformative variable elimination (UVE) were used to select the most important variables for prediction. The average decrements of 23.20 % for root mean square error of cross-validation (RMSECV) and 64.90 % for the absolute value between root mean square error of calibration and RMSECV (AV_RMSE) in all 12 cases achieved after over 90 % variables were eliminated. UVE was also found to be helpful to improve the efficiency of SPA’s variable selection in 8/12 cases. The best predictions for EPA, DHA, and DPA were all achieved by NMR spectroscopy (determination coefficients of cross-validation (r CV 2 ) of 0.970, 0.982, and 0.983 and the RMSECV of 11.48, 4.73, and 0.77 mg/g for the EPA, DHA, and DPA predictions, respectively). LNIR spectra also did good predictions similar to NMR. The results demonstrated that the laborious and time-consuming gas chromatography method could be replaced by spectroscopic techniques in tandem with PLSR modeling and variable selection in order to provide a rapid and reliable inspection of PUFA in fish oil.

Mapping the Spatial Distribution of Winter Crops at Sub-Pixel Level Using AVHRR NDVI Time Series and Neural Nets

For large areas, it is difficult to assess the spatial distribution and inter-annual variation of crop acreages through field surveys. Such information, however, is of great value for governments, land managers, planning authorities, commodity traders and environmental scientists. Time series of coarse resolution imagery offer the advantage of global coverage at low costs, and are therefore suitable for large-scale crop type mapping. Due to their coarse spatial resolution, however, the problem of mixed pixels has to be addressed. Traditional hard classification approaches cannot be applied because of sub-pixel heterogeneity. We evaluate neural networks as a modeling tool for sub-pixel crop acreage estimation. The proposed methodology is based on the assumption that different cover type proportions within coarse pixels prompt changes in time profiles of remotely sensed vegetation indices like the Normalized Difference Vegetation Index (NDVI). Neural networks can learn the relation between temporal NDVI signatures and the sought crop acreage information. This learning step permits a non-linear unmixing of the temporal information provided by coarse resolution satellite sensors. For assessing the feasibility and accuracy of the approach, a study region in central Italy (Tuscany) was selected. The task consisted of mapping the spatial distribution of winter crops abundances within 1 km AVHRR pixels between 1988 and 2001. Reference crop acreage information for network training and validation was derived from high resolution Thematic Mapper/Enhanced Thematic Mapper (TM/ETM+) images and official agricultural statistics. Encouraging results were obtained demonstrating the potential of the proposed approach. For example, the spatial distribution of winter crop acreage at sub-pixel level was mapped with a cross-validated coefficient of determination of 0.8 with respect to the reference information from high resolution imagery. For the eight years for which reference information was available, the root mean squared error (RMSE) of winter crop acreage at sub-pixel level was 10%. When combined with current and future sensors, such as MODIS and Sentinel-3, the unmixing of AVHRR data can help in the building of an extended time series of crop distributions and cropping patterns dating back to the 80s.

Oil accumulation in intact olive fruits measured by near infrared spectroscopy–acousto‐optically tunable filter

A field experiment was conducted to test the reliability of the near infrared spectroscopy (NIR)-acousto-optically tunable filter (AOTF) method to measure mesocarp oil content in vivo against nuclear magnetic resonance (NMR) determinations using three different olive cultivars at different stages of ripening. In the partial least squares model carried out for the cultivar 'Arbequina', the coefficient of determination in calibration (R(2)c) was 0.991, while the coefficient of determination in cross-validation (R(2)cv) was 0.979. For the cultivar 'Frantoio' the indexes were 0.982 and 0.971, respectively; while for the cultivar 'Leccino' R(2)c was 0.977 and R(2)cv was 0.965. Finally, for the combined model (sum of the three varieties) these indexes were 0.921 and 0.903, respectively. The residual predictive deviation (RPD) ratio was insufficient for the predictive model of cultivar 'Leccino' only (1.98), whereas in the other cases the RPD ratios were completely sufficient, within the estimation range over 2.5-3 (2.61 in the global model, and 4.23 in the cultivar 'Frantoio'), or in describing a large capacity with values greater than 5, as in the cultivar 'Arbequina' (9.58). NIR-AOTF spectroscopy proved to be a novel, rapid and reliable method to monitor the oil accumulation process in intact olive fruits in the field. The innovative approach of coupling NIR and NMR technologies opens up new scenarios for determining the optimal time for harvesting olive trees to obtain maximum oil production.

Rapid determination of swine available energy and amino acids in corn distillers dried grains with solubles by near-infrared reflectance spectroscopy

The possibility of using near-infrared reflectance spectroscopy (NIRS) for quantitative determination of available energy (calculated digestible energy and metabolizable energy for swine) and essential amino acids in corn DDGS was investigated in this study. Ninety-three samples were collected from 18 plants in China over a three year period and randomly split into a calibration set (N=70) and a validation set (N=23). Calibrations were performed by modified partial least squared algorithm and 15 different spectral pretreatments. The results showed that the second derivative mathematical treatment gave the best prediction performance for most constituents. The coefficient of determination for cross-validation (RCV2) and validation (Rval2) were excellent for available energy (RCV2 ranged from 0.93 to 0.94, Rval2 ranged from 0.88 to 0.91) and most essential amino acids (RCV2 ranged from 0.85 to 0.93, Rval2 ranged from 0.88 to 0.95) with the exception of arginine, lysine, tryptophan and valine. The prediction performance measuring amino acids by NIRS was better than that obtained by protein regression. These results indicate that NIRS can be used as a quantitative method for the rapid determination of energy and most amino acids in corn DDGS.

Visible and Near-Infrared Diffuse Reflectance Spectroscopy for Prediction of Soil Properties near a Copper Smelter

Spatial and temporal monitoring of soil properties in smelting regions requires collection of a large number of samples followed by laboratory cumbersome and time-consuming measurements. Visible and near-infrared diffuse reflectance spectroscopy (VNIR-DRS) provides a rapid and inexpensive tool to predict various soil properties simultaneously. This study evaluated the suitability of VNIR-DRS for predicting soil properties, including organic matter (OM), pH, and heavy metals (Cu, Pb, Zn, Cd, and Fe), using a total of 254 samples collected in soil profiles near a large copper smelter in China. Partial least square regression (PLSR) with cross-validation was used to relate soil property data to the reflectance spectral data by applying different preprocessing strategies. The performance of VNIR-DRS calibration models was evaluated using the coefficient of determination in cross-validation (R2cv) and the ratio of standard deviation to the root mean standard error of cross-validation (SD/RMSEcv). The models provided fairly accurate predictions for OM and Fe (R2cv > 0.80, SD/RMSEcv > 2.00), less accurate but acceptable for screening purposes for pH, Cu, Pb, and Cd (0.50 < R2cv < 0.80, 1.40 < SD/RMSEcv < 2.00), and poor accuracy for Zn (R2CV < 0.50, SD/RMSEcv < 1.40). Because soil properties in contaminated areas generally show large variation, a comparative large number of calibrating samples, which are variable enough and uniformly distributed, are necessary to create more accurate and robust VNIR-DRS calibration models. This study indicated that VNIR-DRS technique combined with continuously enriched soil spectral library could be a nondestructive alternative for soil environment monitoring.

基于湿地植物光谱的水体总氮估测

利用再生水补充城市湿地是目前湿地恢复与重建的主要方向,然而再水中高浓度的氮、磷含量极易导致水体富营养化。遥感技术已成为富营养化监测的重要手段,但对于植被覆盖水域的富营养化直接探测存在一定的局限性。以北京市典型再生水补水湿地奥林匹克公园南园湿地为研究区,利用湿地植物光谱进行水体富营养化主控因子总氮的遥感探测。测定芦苇(<em>Phragmites australis</em>)和香蒲(<em>Typha angustifolia</em>)的叶片光谱及水体总氮含量,在对数据进行预处理的基础上建立二者的关系模型,包括单变量模型(比值光谱指数(SR)模型和归一化差值光谱指数(ND)模型),与多变量模型(逐步多元线性回归(SMLR)模型和偏最小二乘回归(PLSR)模型),并利用交叉验证决定系数(<em>R</em>_cv²)和均方根误差(RMSE_cv)进行模型精度检验。结果表明,不同回归模型相比,多变量回归模型精度较高;多变量回归模型中,PLSR模型精度较高,<em>R</em>_cv²可达0.72,RMSE_cv仅为0.24,是建立湿地植物光谱与水体总氮含量关系的最优模型。不同湿地植物类型相比,利用芦苇反射光谱建立的各种预测模型的精度都高于香蒲。其他环境因子(总磷)也是影响TN含量与湿地植物反射光谱关系的重要因素。研究成果可以弥补现有水体富营养化遥感探测的不足,并为再生水利用的城市湿地水质监测与管理提供有力的科学依据。;Supplying urban wetlands with reclaimed water is recognized as a superior way for wetland restoration and reconstruction. However, the high concentration of nitrogen and phosphorus in reclaimed water can easily lead to water eutrophication. Although remote sensing technology has become a useful tool to monitor the eutrophication of water body, it is usually employed to detect eutrophication in open water. Limited applications have been found in measuring eutrophication of wetland covered by vegetation. Utilizing plants spectral response to environment can monitor environmental changes. This study explores the possibility to use wetland vegetation reflectance spectra in estimating total nitrogen content which is one of the key indicators of water eutrophication. The South Wetland in the Olympic Park in Beijing, a typical wetland using reused water, was selected as our study area. The leaf reflectance spectra of main wetland plants, reed (<em>Phragmites australis</em>) and cattail (<em>Typha angustifolia</em>), were acquired by means of an ASD FieldSpec 3 spectrometer (350-2500nm). Water quality samples were collected at the same time and analyzed by Center for Environmental Quality Test, Tsinghua University subsequently. The research established several univariate models including simple ratio spectral index (SR) model and normalized difference spectral index (ND) model, as well as multivariate models including stepwise multiple linear regression (SMLR) model and partial least squares regression (PLSR) model. The accuracy of these models was tested through cross-validated coefficient of determination (<em>R</em>_cv²) and cross-validated root mean square error (RMSE_cv). The results have shown that 1) In comparison with univariate techniques, multivariate regressions can improve the estimation of total nitrogen concentration in water. The accuracy of PLSR model was the highest (<em>R</em>_cv²=0.72, RMSE_cv=0.24) among all models. PLSR provides the most useful explorative tool for unraveling the relationship between spectral reflectance of wetland plants and total nitrogen content in water at leaf scale. 2) The accuracy of prediction models built in this study using <em>Phragmites australis</em> reflectance spectra is higher than those using <em>Typha angustifolia</em> reflectance spectra. 3) Other environmental factors should also be discreetly considered in modeling exercise. Total phosphorus is found to have impact on the relationship between TN and reflectance spectra from wetland vegetation. Strong predictive power for multiple regression equations has been achieved when the range of total phosphorus was restricted. The result from this study can not only fill the gaps in the detection of eutrophication using remote sensing, but also provide a strong scientific basis for the water quality monitoring and management of urban wetlands using recycled water.

Mid-infrared prediction of lactoferrin content in bovine milk: potential indicator of mastitis

Lactoferrin (LTF) is a milk glycoprotein favorably associated with the immune system of dairy cows. Somatic cell count is often used as an indicator of mastitis in dairy cows, but knowledge on the milk LTF content could aid in mastitis detection. An inexpensive, rapid and robust method to predict milk LTF is required. The aim of this study was to develop an equation to quantify the LTF content in bovine milk using mid-infrared (MIR) spectrometry. LTF was quantified by enzyme-linked immunosorbent assay (ELISA), and all milk samples were analyzed by MIR. After discarding samples with a coefficient of variation between 2 ELISA measurements of more than 5% and the spectral outliers, the calibration set consisted of 2499 samples from Belgium (n = 110), Ireland (n = 1658) and Scotland (n = 731). Six statistical methods were evaluated to develop the LTF equation. The best method yielded a cross-validation coefficient of determination for LTF of 0.71 and a cross-validation standard error of 50.55 mg/l of milk. An external validation was undertaken using an additional dataset containing 274 Walloon samples. The validation coefficient of determination was 0.60. To assess the usefulness of the MIR predicted LTF, four logistic regressions using somatic cell score (SCS) and MIR LTF were developed to predict the presence of mastitis. The dataset used to build the logistic regressions consisted of 275 mastitis records and 13 507 MIR data collected in 18 Walloon herds. The LTF and the interaction SCS × LTF effects were significant (P < 0.001 and P = 0.02, respectively). When only the predicted LTF was included in the model, the prediction of the presence of mastitis was not accurate despite a moderate correlation between SCS and LTF (r = 0.54). The specificity and the sensitivity of models were assessed using Walloon data (i.e. internal validation) and data collected from a research herd at the University of Wisconsin – Madison (i.e. 5886 Wisconsin MIR records related to 93 mastistis events – external validation). Model specificity was better when LTF was included in the regression along with SCS when compared with SCS alone. Correct classification of non-mastitis records was 95.44% and 92.05% from Wisconsin and Walloon data, respectively. The same conclusion was formulated from the Hosmer and Lemeshow test. In conclusion, this study confirms the possibility to quantify an LTF indicator from milk MIR spectra. It suggests the usefulness of this indicator associated to SCS to detect the presence of mastitis. Moreover, the knowledge of milk LTF could also improve the milk nutritional quality.

Use of near infrared spectroscopy for intramuscular fat selection in rabbits

The potential use of near infrared spectroscopy (NIRS) for the determination of intramuscular fat (IMF) content in rabbit selection programmes was evaluated. One hundred and thirty seven rabbits from three different synthetic lines slaughtered between 5 and 61 weeks of age were used for NIR calibration. Longissimus muscles (LM) were homogenised, freeze-dried and scanned by NIRS reflectance and total lipid content was chemically analysed. Calibration equation parameters reported appropriate results for IMF (cross-validation standard error, SECV=0.07g/100g muscle; cross-validation coefficient of determination, R2=0.98 and residual predictive deviation of cross-validation, RPD=7.57). Another 88 rabbits were used to study the suitability of NIR spectroscopy in selection programmes. Intramuscular fat was measured in LM using chemical and NIRS analyses. Descriptive statistics showed that NIRS could be a proper technique to average comparison, but regression analyses (R2=0.92) and rank correlation measures, especially Kendall's tau-b correlation coefficient (0.83), indicated that NIRS may not be accurate enough to predict individual genetic values and produce ranking of animals. However, NIRS technique could be applied in truncated selection where the efficiency of the method is measured by the response to selection. Selection can be done on second parities using the IMF value of two full sibs of first parities. Ten females and 5 males can be selected as parents to establish a new population of 40 females and 5 males. The IMF values were similar between animals selected on the basis of chemically-determined IMF and NIRS-predicted IMF content. Results of the experiment confirmed the potential of NIRS for the determination of IMF content in rabbit selection programmes instead of using laborious chemical methods

Using FTIR spectroscopy to model alkaline pretreatment and enzymatic saccharification of six lignocellulosic biomasses

Fourier transform infrared, attenuated total reflectance (FTIR-ATR) spectroscopy, combined with partial least squares (PLS) regression, accurately predicted solubilization of plant cell wall constituents and NaOH consumption through pretreatment, and overall sugar productions from combined pretreatment and enzymatic hydrolysis. PLS regression models were constructed by correlating FTIR spectra of six raw biomasses (two switchgrass cultivars, big bluestem grass, a low-impact, high-diversity mixture of prairie biomasses, mixed hardwood, and corn stover), plus alkali loading in pretreatment, to nine dependent variables: glucose, xylose, lignin, and total solids solubilized in pretreatment; NaOH consumed in pretreatment; and overall glucose and xylose conversions and yields from combined pretreatment and enzymatic hydrolysis. PLS models predicted the dependent variables with the following values of coefficient of determination for cross-validation (Q²): 0.86 for glucose, 0.90 for xylose, 0.79 for lignin, and 0.85 for total solids solubilized in pretreatment; 0.83 for alkali consumption; 0.93 for glucose conversion, 0.94 for xylose conversion, and 0.88 for glucose and xylose yields. The sugar yield models are noteworthy for their ability to predict overall saccharification through combined pretreatment and enzymatic hydrolysis per mass dry untreated solids without a priori knowledge of the composition of solids. All wavenumbers with significant variable-important-for-projection (VIP) scores have been attributed to chemical features of lignocellulose, demonstrating the models were based on real chemical information. These models suggest that PLS regression can be applied to FTIR-ATR spectra of raw biomasses to rapidly predict effects of pretreatment on solids and on subsequent enzymatic hydrolysis.

Quantitative structure–activity relationship analysis of acute toxicity of diverse chemicals to Daphnia magna with whole molecule descriptors

Quantitative structure–activity relationship analysis and estimation of toxicological effects at lower-mid trophic levels provide first aid means to understand the toxicity of chemicals. Daphnia magna serves as a good starting point for such toxicity studies and is also recognized for regulatory use in estimating the risk of chemicals. The ECOTOX database was queried and analysed for available data and a homogenous subset of 253 compounds for the endpoint LC50 48 h was established. A four-parameter quantitative structure–activity relationship was derived (coefficient of determination, r 2 = 0.740) for half of the compounds and internally validated (leave-one-out cross-validated coefficient of determination, = 0.714; leave-many-out coefficient of determination, = 0.738). External validation was carried out with the remaining half of the compounds (coefficient of determination for external validation, = 0.634). Two of the descriptors in the model (log P, average bonding information content) capture the structural characteristics describing penetration through bio-membranes. Another two descriptors (energy of highest occupied molecular orbital, weighted partial negative surface area) capture the electronic structural characteristics describing the interaction between the chemical and its hypothetic target in the cell. The applicability domain was subsequently analysed and discussed.

Using FTIR to predict saccharification from enzymatic hydrolysis of alkali‐pretreated biomasses

Fourier transform infrared, attenuated total reflectance (FTIR-ATR) spectroscopy combined with partial least squares (PLS) regression accurately predicted 72-h glucose and xylose conversions (g sugars/100 g potential sugars) and yields (g sugars/100 g dry solids) from cellulase-mediated hydrolysis of alkali-pretreated lignocellulose. Six plant biomasses that represent a variety of potential biofuel feedstocks--two switchgrass cultivars, big bluestem grass, a low-impact, high-diversity mixture of 32 species of prairie biomasses, mixed hardwood, and corn stover--were subjected to four levels of low-temperature NaOH pretreatment to produce 24 samples with a wide range of potential digestibility. PLS models were constructed by correlating FTIR spectra of pretreated samples to measured values of gluose and xylose conversions and yields. Variable selection, based on 90% confidence intervals of regression-coefficient matrices, improved the predictive ability of the models, while simplifying them considerably. Final models predicted sugar conversions with coefficient of determination for cross-validation (Q(2)) values of 0.90 for glucose and 0.89 for xylose, and sugar yields with Q(2) values of 0.92 for glucose and 0.91 for xylose. The sugar-yield models are noteworthy for their ability to predict enzymatic saccharification per mass dry solids without a priori knowledge of the composition of the solids. All peaks retained in the final regression coefficient matrices were previously assigned to chemical bonds and functional groups in lignocellulose, demonstrating that the models were based on real chemical information. This study demonstrates that FTIR spectroscopy combined with PLS regression can be used to rapidly estimate sugar conversions and yields from enzymatic hydrolysis of pretreated plant biomass.

Estimating the spatial distribution of green herbage biomass and quality by geostatistical analysis with field hyperspectral measurements

This study aimed to estimate green herbage biomass (GBM) and crude protein (CP) concentrations of a mixed-sown pasture in Hokkaido, Japan using ground based hyperspectral measurements and geostatistical analysis. The mixed-sown pasture consisted of a relatively flat section renovated by over-seeding a grass (Subunit 1, 2.6 ha) and a hilly aged section (Subunit 2, 5.0 ha). Hyperspectral reflectance and plant data were collected for 22 days in August 2009 from 88 plots within the two subunits. For mapping, separate spectral readings, without plant sampling, were obtained from a total of 347 plots along permanent transects in the pasture. Genetic algorithm-based wavebands selection with partial least squares (GA-PLS) regression analyses was performed to predict GBM and CP concentrations using both reflectance and first derivative reflectance (FDR) datasets. Then, geostatistical analysis with semivariograms was conducted to determine sampling interval of GBM and CP concentration in Subunits 1 and 2. In the GA-PLS analysis, the most accurate results were obtained by calibration of GBM using FDR (cross-validated coefficient of determination, ; cross-validated root mean square error, RMSECV = 42) and of the CP concentration using raw reflectance (, RMSECV = 2.02). Geostatistical analysis with semivariograms showed that, at the landscape scale, the GBM patch sizes in Subunits 1 and 2 were 31 and 67 m, respectively, and those of the CP concentration were 37 and 54 m. These values indicate that the spatial distribution patterns of these pasture parameters were more heterogeneous in Subunit 1 than in Subunit 2. The analysis result also indicates that the sampling interval for GBM and CP concentration should be <15 m. This work shows the ability to estimate the spatial distribution of GBM and CP concentration for the implementation of site-specific grazing management.