Variable Importance In The Projection Research Articles

Significant amino acids in aroma compound profiling during yeast fermentation analyzed by PLS regression

The biochemical formation of aroma-active metabolites determines different characteristics of aroma and taste in fermented beverages. Amino acid utilization by yeast during brewer's wort fermentation is seen as linked to flavor profile. For major understanding of the relationship between amino acid composition in wort and their impact on the synthesis of aroma-active metabolites amino acids were varied in synthetic medium and fermented by Saccharomyces cerevisiae strain S81 and Saccharomyces pastorianus var. carlsbergensis strain S23, respectively. After fermentation of different amino acid combinations higher alcohols and esters were detected by gas chromatography to determine specific aroma compound spectrum. Partial least square (PLS) regression and variable importance in the projection (VIP) were used to establish a relation between amino acids and the resulting concentrations of aroma compounds by means of pattern recognition, indicating most representative amino acids in aroma compound synthesis during fermentation. Thus, a fingerprint of amino acid importance on the detected aroma compound spectrum was created. The most important explanatory variables affecting the synthesis of aroma-active substances of S. pastorianus var. carlsbergensis are leucine, isoleucine, valine, glutamine, cysteine and surprisingly proline. In case of S. cerevisiae the significant amino acids are leucine, isoleucine, valine, histidine, glutamine and again proline.

Predictive-property-ranked variable reduction in partial least squares modelling with final complexity adapted models: Comparison of properties for ranking

The calibration performance of partial least squares regression for one response (PLS1) can be improved by eliminating uninformative variables. Many variable-reduction methods are based on so-called predictor-variable properties or predictive properties, which are functions of various PLS-model parameters, and which may change during the steps of the variable-reduction process. Recently, a new predictive-property-ranked variable reduction method with final complexity adapted models, denoted as PPRVR-FCAM or simply FCAM, was introduced. It is a backward variable elimination method applied on the predictive-property-ranked variables. The variable number is first reduced, with constant PLS1 model complexity A, until A variables remain, followed by a further decrease in PLS complexity, allowing the final selection of small numbers of variables.In this study for three data sets the utility and effectiveness of six individual and nine combined predictor-variable properties are investigated, when used in the FCAM method. The individual properties include the absolute value of the PLS1 regression coefficient (REG), the significance of the PLS1 regression coefficient (SIG), the norm of the loading weight (NLW) vector, the variable importance in the projection (VIP), the selectivity ratio (SR), and the squared correlation coefficient of a predictor variable with the response y (COR). The selective and predictive performances of the models resulting from the use of these properties are statistically compared using the one-tailed Wilcoxon signed rank test.The results indicate that the models, resulting from variable reduction with the FCAM method, using individual or combined properties, have similar or better predictive abilities than the full spectrum models. After mean-centring of the data, REG and SIG, provide low numbers of informative variables, with a meaning relevant to the response, and lower than the other individual properties, while the predictive abilities are similar or better. SIG has the best selective ability of all individual and combined properties, while the predictive ability is similar. REG is faster than SIG. This means that variable reduction with the FCAM method is preferably conducted with properties REG or SIG. The selective ability of REG can be improved by combining it with NLW or VIP.

PP098. Lipidic fingerprinting in women with early-onset preeclampsia: A first look

preeclampsia is characterized by intense inflammatory response and an anti-angiogenic state. Maternal obesity has been considered to have important impact on the genesis of preeclampsia as lipotoxicity leads to maternal endothelial dysfunction and chronic inflammation. Here we investigate the plasma lipid profile of preeclamptic women. identify possible lipid biomarkers for preeclampsia. this study included 8 pregnant women with early-onset preeclampsia (before 34weeks gestation) and 8 normal pregnant women. Each patient in the preeclampsia group was matched to a patient in the control group according to gestational age at the time of sample collection. All patients in the control group were followed until term and had normal outcomes. To investigate the lipid profile, lipids were extracted from plasma samples using the Bligh-Dyer protocol and the extracts were subjected to MALDI-TOF Mass Spectrometry. Data matrix was exported for partial least squares discriminant analysis. All the variables analysed were sorted by a score number named Variable Importance in the Projection. The major discriminant variables were selected and underwent to Mann-Whitney U test. a total of 1290 ions were initially identified during lipidomic assessment. Twelve m/z signals were highlighted as the most important lipids for the discrimination of patients with preeclampsia. The identification of these differential lipids was carried out through Lipid Database Search. The main classes identified were Glycerophosphocholines [GP01], Glycerophosphoserines [GP03], Glycerophosphoglycerols [GP04], Glycosyldiradylglycerols [GL05] and Glycerophosphates [GP10]. Our results suggest that some lipid species may be potential biomarkers for early-onset preeclampsia.

Metabonomic study of intervention effects of Morinda officinalis on ‘Kidney-yang deficiency syndrome’

To investigate the metabolic profile of hydrocortisone-induced 'Kidney-yang deficiency syndrome'in rats and the intervention effects of Morinda officinalis. Proton nuclear magnetic resonance (1H-NMR) technique was used to analyze the rat metabonome in serum. Orthogonal partial least squares discriminant analysis (OPLS-DA) were processed to analyze the metabonome difference between the control and hydrocortisone treated samples. Twelve potential biomarkers were selected, via the parameter of variable importance in the projection (VIP). Principal components analysis (PCA) was employed to process the data from the M. officinalis. treatment group and the intervention effects of M. officinalis, was investigated through the selected potential biomarkers. After hydrocortisone treatment, the energy metabolism, amino acids metabolism and gut microflora environment were seriously disturbed and transmethylation was surpressed. M. officinalis could effectively alleviate the disturbance of energy and amino acids metabolism and enhance transmethylation, but could not modulate the gut microflora environment. The results obtained suggested that metabonomic studies could better reflect the whole status of metabolism in bio-systems, and could be treated as a potential powerful approach in pharmacological studies and investigation of the essence of 'syndrome' in traditional Chinese medicine.

Combining bootstrap and uninformative variable elimination: Chemometric identification of metabonomic biomarkers by nonparametric analysis of discriminant partial least squares

Interpretation and mining of complex metabonomic data depend heavily on proper use of chemometric methods. Due to the “small n” paradigm and the absence of sufficient information concerning distribution of data, the classical parametric methods based on known theoretical distributions are sometimes unsuitable or unreliable to treat such data. Therefore, nonparametric methods requiring no or very limited assumptions provide useful alternative tools in many practical applications. In this paper, a new discriminant partial least squares combined with bootstrap and uninformative variable elimination (DPLS–BS–UVE) method is proposed for biomarker discovery in metabonomics. The method was tested on two real chromatographic data sets containing plasma metabolic profilings for S180 and H22 tumor-bearing mice. A robust version of cj was used as the cutoff criterion. The results of biomarker discovery were compared with those obtained using variable importance in the projection (VIP) as well as BS. It is demonstrated that similar results are obtained using the three methods and DPLS–BS–UVE could provide easy interpretation of raw data. When the resampling unit increases to 500, the results were not significantly affected. In conclusion, DPLS–BS–UVE is a reliable alternative method for biomarker discovery, especially when the sample size is small.

Partial Least Squares Regression for analyzing walnut phenology in California

Many biological processes produce only one quantitative outcome per year, resulting from temperatures and precipitation during hundreds of days leading up to the event. Traditional regression approaches incur problems in such a setting, because independent variables are highly autocorrelated and their number often greatly exceeds the number of observations. Partial Least Squares Regression (PLS), a statistical analysis tool developed to handle these situations and widely used in hyperspectral remote sensing, was tested for its usefulness for explaining the climate responses of biological processes, using walnut phenology in California as an example.Observations of first female bloom, first male bloom and leaf emergence of three walnut cultivars at Davis, CA were coupled with daily temperature data since 1951. The dataset was analyzed by PLS, using three temperature inputs: (1) daily mean temperatures, (2) 11-day running means of daily mean temperatures and (3) monthly mean temperatures. For all data constellations, the Variable-Importance-in-the-Projection (VIP) statistic indicated a number of periods, during which temperatures were important determinants of phenological events, and the model-coefficients-of-the-centered-and-scaled-data (MC) statistic showed the direction, in which high temperatures during these phases influenced walnut flowering and leaf emergence. In all analyses, a delaying effect of warm winters, and an advancing effect of warm springs were clearly visible. It was also possible to identify the transition between the chilling and forcing phases, and the VIP and MC plots indicated quantitative differences in the effectiveness of winter chill during different phases of the dormancy season. Such effects have not been captured in any phenology models currently applied to fruit trees, indicating that PLS has potential to help refine such models. PLS can also be used for guiding experimental research by pinpointing the parts of the season that are most important for the timing of budburst. Results suggested that more than 20 years of observed data were necessary for producing clearly recognizable temperature response patterns, limiting the applicability of PLS to long time series.

Label-free quantitative proteomic analysis reveals dysfunction of complement pathway in peripheral blood of schizophrenia patients: evidence for the immune hypothesis of schizophrenia

Schizophrenia is a complex mental disease caused by a combination of serial alterations in genetic and environmental factors. Although the brain is usually considered as the most relevant organ in schizophrenia, accumulated evidence suggests that peripheral tissues also contribute to this disease. In particular, abnormalities of the immune system have been identified in the peripheral blood of schizophrenia patients. To screen the serum proteomic signature of schizophrenia patients, we conducted shotgun proteomic analysis on serum samples of schizophrenia patients and healthy controls. High-abundance proteins were eliminated by immunoaffinity before LC-MS/MS analysis. The multivariate statistical test partial least squares-discriminant analysis (PLS-DA) was applied to build models for screening out variable importance in the projection (VIP) and 27 proteins were identified as being responsible for discriminating between the proteomic profiles of schizophrenia patients and healthy controls. Pathway analysis based on these 27 proteins revealed that complement and coagulation cascades was the most significant pathway. ELISA-based activity analyses indicated that the alternative complement pathway was suppressed in schizophrenia patients. Ingenuity pathways analysis was used to conduct the interaction network of 27 proteins. The network exhibited common features such as, nervous system development and function, humoral immune response and inflammatory response, and highlighted some proteins with important roles in the immune system, such as hub nodes. Our findings indicate dysregulation of the alternative complement pathway in schizophrenia patients. The protein interaction network enhances the interpretation of proteomic data and provides evidence that the immune system may contribute to schizophrenia.

Serum metabolic profiling of human gastric cancer based on gas chromatography/mass spectrometry

Research on molecular mechanisms of carcinogenesis plays an important role in diagnosing and treating gastric cancer. Metabolic profiling may offer the opportunity to understand the molecular mechanism of carcinogenesis and help to non-invasively identify the potential biomarkers for the early diagnosis of human gastric cancer. The aims of this study were to explore the underlying metabolic mechanisms of gastric cancer and to identify biomarkers associated with morbidity. Gas chromatography/mass spectrometry (GC/MS) was used to analyze the serum metabolites of 30 Chinese gastric cancer patients and 30 healthy controls. Diagnostic models for gastric cancer were constructed using orthogonal partial least squares discriminant analysis (OPLS-DA). Acquired metabolomic data were analyzed by the nonparametric Wilcoxon test to find serum metabolic biomarkers for gastric cancer. The OPLS-DA model showed adequate discrimination between cancer and non-cancer cohorts while the model failed to discriminate different pathological stages (I-IV) of gastric cancer patients. A total of 44 endogenous metabolites such as amino acids, organic acids, carbohydrates, fatty acids, and steroids were detected, of which 18 differential metabolites were identified with significant differences. A total of 13 variables were obtained for their greatest contribution in the discriminating OPLS-DA model [variable importance in the projection (VIP) value >1.0], among which 11 metabolites were identified using both VIP values (VIP >1) and the Wilcoxon test. These metabolites potentially revealed perturbations of glycolysis and of amino acid, fatty acid, cholesterol, and nucleotide metabolism of gastric cancer patients. These results suggest that gastric cancer serum metabolic profiling has great potential in detecting this disease and helping to understand its metabolic mechanisms.

PLS discriminant analysis applied to conventional sensory profiling data

Several methods have been proposed in the literature to analyze conventional sensory profiling data. We focus on factor analytical methods which have been extensively used due to their ability to produce graphical displays which are both useful and easy to interpret. Available factor analytical methods include principal components analysis on averaged assessors data or on the data matrix obtained by stacking the assessors’ datasets one on top of the others, and canonical variates analysis; each method has advantages and drawbacks. As an alternative it is advocated the use of PLS discriminant analysis, which is at the intersection of the methods mentioned above. This method of analysis takes account of the within and between product variations while minimizing the impact of multicolinearity. It provides statistical tools to assess on the one hand the agreement among assessors and the discrimination among products by means of the between to total variance ratio, and on the other hand the relative importance of variables by means of VIP (variable importance in the projection) indices. The VIP indices may also be useful to guide the selection of a subset of relevant attributes from the complete set of attributes. In this paper, PLS discriminant analysis is compared with other methods, and results are illustrated through a case study. In particular, the stability of the various methods is investigated using assessors’ re-sampling (bootstrap) and confidence ellipses.

Data-driven prediction model of indoor air quality in an underground space

Several data-driven prediction methods based on multiple linear regression (MLR), neural network (NN), and recurrent neural network (RNN) for the indoor air quality in a subway station are developed and compared. The RNN model can predict the air pollutant concentrations at a platform of a subway station by adding the previous temporal information of the pollutants on yesterday to the model. To optimize the prediction model, the variable importance in the projection (VIP) of the partial least squares (PLS) is used to select key input variables as a preprocessing step. The prediction models are applied to a real indoor air quality dataset from telemonitoring systems data (TMS), which exhibits some nonlinear dynamic behaviors show that the selected key variables have strong influence on the prediction performances of the models. It demonstrates that the RNN model has the ability to model the nonlinear and dynamic system, and the predicted result of the RNN model gives better modeling performance and higher interpretability than other data-driven prediction models.

Robust weighted regression as a downscaling tool in temperature projections

Downscaling models are developed using robust version of locally weighted regression smoothing scatter plots technique (LOWESS) regression approach for obtaining projections of mean monthly maximum and minimum temperatures (Tmax and Tmin) to Pichola watershed in an arid region in India. Variable Importance in the Projection (VIP) score from Partial Least Squares (PLSs) regression is used to select the variables. A comparison is also done with LOWESS regression approach. The results show that an increasing trend is observed for Tmax and Tmin for A1B, A2 and Bl scenarios whereas no trend is discerned with the COMMIT.

Predictive QSPR modeling of the acidic dissociation constant (pKa) of phenols in different solvents

AbstractGiven the importance of ionization constant (pKa) of phenols in explaining the mechanism of their toxicity, it is of interest to develop theoretical models for the prediction of pKa values of phenols in different solvent systems. In the present communication, we developed predictive QSPR models for pKa values of substituted phenols in seven different solvent systems such as water, dimethyl sulfoxide (DMSO), methanol, dimethylformamide (DMF), acetonitrile (AN), isopropanol, and tert‐butanol using quantum topological molecular similarity (QTMS) descriptors. The data set was divided into training and test sets, and models were developed using partial least squares (PLS) regression from the training set. The predictive potential of the developed models was assessed by the prediction of pKa values of the test set compounds. Root mean square error of prediction (RMSEP) values were used as objective function for selection of the best models in different solvent systems. Good predictive models were developed in all solvent systems except isopropanol. Considering all seven solvent systems, distance descriptors give consistently good results whereas ellipticity descriptors are of less importance. Moreover, plots of ‘variable importance in the projection’ (VIP) for the best models highlight the importance of the bond connecting the phenolic oxygen to the aromatic ring. This suggests the diagnostic nature of QTMS descriptors in identifying the reaction center in acidic dissociation of phenols. Copyright © 2008 John Wiley & Sons, Ltd.

Variable selection in near infrared spectra for the biological characterization of soil and earthworm casts

Near infrared reflectance spectroscopy (NIRS) was used to predict six biological properties of soil and earthworm casts including extracellular soil enzymes, microbial carbon, potential nitrification and denitrification. Partial least squares regression (PLSR) models were developed with a selection of the most important near infrared wavelengths. They reached coefficients of determination ranging from 0.81 to 0.91 and ratios of performance-to-deviation above 2.3. Variable selection with the variable importance in the projection (VIP) method increased dramatically the prediction performance of all models with an important contribution from the 1750–2500 nm region. We discuss whether selected wavelengths can be attributed to macronutrient availability or to microbial biomass. Wavelength selection in NIR spectra is recommended for improving PLSR models in soil research.

NMR and Chemometric methods: A powerful combination for characterization of Balsamic and Traditional Balsamic Vinegar of Modena

This work presents the capability of NMR spectroscopy combined with Chemometrics in predicting the ageing of Balsamic and Traditional Balsamic Vinegar of Modena. The need of an analytical method is an important requirement for both research oriented and commercial evaluation of these very valuable products. 1H NMR spectroscopy, based on the advantage of rapid sample analysis without any manipulation or derivatization, is here proposed as a valid tool to describe Balsamic and Traditional Balsamic Vinegar of Modena. For this purpose, 72 reliable samples, were divided into three different groups according to their ageing process: young (<12 years), old (>12 and <25 years) and extra old (>25 years). Hierarchical Projection to Latent Structures Discriminant Analysis (PLS-DA) allowed us to characterize the ageing process. Variables showing the largest VIP (Variable Importance in the Projection) were extracted from PLS-DA model, thus shedding lights onto the role played by specific compounds in this complex ageing process. Two robust classification models, were built by PLS-DA and Naïve Bayes classifier and compared to prove the accuracy of the representation on both training and test sets. The predictions obtained for 41 “unknown” vinegar samples with these both methods gave more than 80% agreement among them.

Classification and Diagnostic Output Prediction of Cancer Using Gene Expression Profiling and Supervised Machine Learning Algorithms

In this paper, a new supervised clustering and classification method is proposed. First, the application of discriminant partial least squares (DPLS) for the selection of a minimum number of key genes is applied on a gene expression microarray data set. Second, supervised hierarchical clustering based on the information of the cancer type is subsequently proposed to find key gene groups and to group the cancer samples into different subclasses. Here, the weights of the genes in the DPLS are proportional to their importance in the determination of the class labels, that is, the variable importance in the projection (VIP) information of the DPLS method. The power of the gene selection method and the proposed supervised hierarchical clustering method is illustrated on a three microarray data sets of leukemia, breast, and colon cancer. Supervised machine learning algorithms thus enable the subtype classification 3 data sets solely on the basis of molecular-level monitoring. Compared to unsupervised clustering, the supervised method performed better for discriminating between cancer types and cancer subtypes for the leukemia data set. The performance of the proposed method, using only a limited set of informative genes, is demonstrated to be comparable or better than results reported in the literature for the three data sets. Furthermore the method was successful in predicting the outcome of medical treatment (success or failure) based on the microarray data, which could make the method an important tool for clinical doctors.

DFT study on the structure-toxicity relationship of dioxin compounds using PLS analysis

Density functional theory (DFT) at B3LYP/6-311G** level was employed to optimise the dioxin compounds, i.e., 25 polychlorinated or brominated dibenzo-p-dioxins (PCDDs or PBDDs) and 34 polychlorinated dibenzofurans (PCDFs) involved in this investigation. Three groups of descriptors mainly related to chemical reactivity, molecular overall charge distribution and thermochemical property were calculated. With partial least squares (PLS) analysis and variable importance in the projection (VIP), the least significant descriptors were removed from the quantitative structure-activity relationship (QSAR), which was focused on exploring the influential factors responsible for the variance of binding affinities of dioxins to aryl hydrocarbon receptor (AhR). With better-improved and predictive QSAR ( = 0.827), further understanding of the nature of toxicity was available. Both dispersion interaction and electrostatic interaction were considered to be important and together capable of accounting for the most part of the total binding affinities, though the former could make more contribution than the latter. Comparatively, the long-range dispersion interaction should be very small.

Structural Requirements of Angiotensin I‐Converting Enzyme Inhibitory Peptides: Quantitative Structure‐Activity Relationship Modeling of Peptides Containing 4‐10 Amino Acid Residues

AbstractModels for angiotensin converting enzyme (ACE) inhibitory peptides with varied lengths are proposed based on results from partial least square regression analysis of the potency of known peptides. Modeling was performed using partial least square regression with inhibitory activity of peptides as the dependent variable (Y) and the physicochemical properties of amino acids as the predictor X matrix. Variable importance in the projection (VIP) analysis of individual amino acid residues at each position revealed that the C‐terminal tetrapeptide residues of long‐chain peptides were more important to their ACE‐inhibitory activity than the C‐terminal tripeptide residues. The most likely preferred amino acid residues starting from C‐terminus are tyrosine and cysteine for the first position, histidine, tryptophan and methionine for the second position with isoleucine, leucine, valine and methionine for the third position, and tryptophan for the fourth position. We concluded that the reported structural requirements of ACE‐inhibitory peptides provide useful information that can be used for the development of more efficacious ACE‐inhibitory peptides.

Performance of some variable selection methods when multicollinearity is present

Variable selection is one of the important practical issues for many scientific engineers. Although the PLS (partial least squares) regression combined with the VIP (variable importance in the projection) scores is often used when the multicollinearity is present among variables, there are few guidelines about its uses as well as its performance. The purpose of this paper is to explore the nature of the VIP method and to compare with other methods through computer simulation experiments. We design 108 experiments where observations are generated from true models considering four factors–the proportion of the number of relevant predictors, the magnitude of correlations between predictors, the structure of regression coefficients, and the magnitude of signal to noise. Confusion matrix is adopted to evaluate the performance of PLS, the Lasso, and stepwise method. We also discuss the proper cutoff value of the VIP method to increase its performance. Some practical hints for the use of the VIP method are given as simulation results.

Deflection of 4o-inch Brick Wall under pressure of one finger can be measured by device recently constructed : (Bureau of Standards)

Invasive plants pose significant threats to biodiversity and ecosystem function globally, leading to costly monitoring and management effort. While remote sensing promises cost-effective, robust and repeatable monitoring tools to support intervention, it has been largely restricted to airborne platforms that have higher spatial and spectral resolutions, but which lack the coverage and versatility of satellite-based platforms. This study tests the ability of the WorldView-2 (WV2) eight-band satellite sensor for detecting the invasive shrub mesquite (Prosopis spp.) in the north-west Pilbara region of Australia. Detectability was challenged by the target taxa being largely defoliated by a leaf-tying biological control agent (Gelechiidae: Evippe sp. #1) and the presence of other shrubs and trees. Variable importance in the projection (VIP) scores identified bands offering greatest capacity for discrimination were those covering the near-infrared, red, and red-edge wavelengths. Wavelengths between 400 nm and 630 nm (coastal blue, blue, green, yellow) were not useful for species level discrimination in this case. Classification accuracy was tested on three band sets (simulated standard multispectral, all bands, and bands with VIP scores ≥1). Overall accuracies were comparable amongst all band-sets (Kappa = 0.71–0.77). However, mesquite omission rates were unacceptably high (21.3%) when using all eight bands relative to the simulated standard multispectral band-set (9.5%) and the band-set informed by VIP scores (11.9%). An incremental cover evaluation on the latter identified most omissions to be for objects <16 m2. Mesquite omissions reduced to 2.6% and overall accuracy significantly improved (Kappa = 0.88) when these objects were left out of the confusion matrix calculations. Very high mapping accuracy of objects >16 m2 allows application for mapping mesquite shrubs and coalesced stands, the former not previously possible, even with 3 m resolution hyperspectral imagery. WV2 imagery offers excellent portability potential for detecting other species where spectral/spatial resolution or coverage has been an impediment. New generation satellite sensors are removing barriers previously preventing widespread adoption of remote sensing technologies in natural resource management.