Multiblock Approach Research Articles

A chemometric approach to the evaluation of the ageing ability of red wines

One of the most important quality attributes of red wines is its ability to withstand and ageing process. The multidimensionality of the ageing ability concept, which includes a combination of colour, taste, mouthfeel, and aroma was reported in previous studies. Phenolic compounds are largely or partially involved in most of these wine attributes and have been proposed as potential candidates to evaluate the ageing ability of red wines. The phenolic and colour properties of a large number of wines were measured during a barrel and bottle ageing process of 24 months. To our understanding, a wine that needs the longest time to reach optimal phenolic quality is considered a wine with higher ability to age (concentration assumption). Moreover, a wine that is able to maintain its optimal phenolic quality for longer will also be a wine with high ageing ability (stability assumption). Based on the formulated assumptions a scoring system was used to identify those wines with theoretically high ability to age. As expected, these wines contained initial high levels of tannins, total phenols and anthocyanins, including high polymeric pigment presence and enhanced colour properties. Interestingly, high ageing ability wines showed a smaller change in the colour properties over time which might be indicating slower pigment formation rates. In addition, a chemometric attempt was undertaken to explore an ageing index based on initial phenolic content. The evolution of the index over time using a multi-block approach showed stability for the index values, in line with the short term stability of tannin and total phenol levels. Finally, promising results were obtained as two thirds of the wines were correctly classified when a validation task between the ageing index and the score values was attempted.

Geographical Classification of Italian Saffron (Crocus sativus L.) by Multi-Block Treatments of UV-Vis and IR Spectroscopic Data.

One-hundred and fourteen samples of saffron harvested in four different Italian areas (three in Central Italy and one in the South) were investigated by IR and UV-Vis spectroscopies. Two different multi-block strategies, Sequential and Orthogonalized Partial Least Squares Linear Discriminant Analysis (SO-PLS-LDA) and Sequential and Orthogonalized Covariance Selection Linear Discriminant Analysis (SO-CovSel-LDA), were used to simultaneously handle the two data blocks and classify samples according to their geographical origin. Both multi-block approaches provided very satisfying results. Each model was investigated in order to understand which spectral variables contribute the most to the discrimination of samples, i.e., to the characterization of saffron harvested in the four different areas. The most accurate solution was provided by SO-PLS-LDA, which only misclassified three test samples over 31 (in external validation).

Simultaneous exploration of nutrients and pollutants in human milk and their impact on preterm infant growth: An integrative cross-platform approach

Early nutritional management including fortified human breastmilk is currently recommended to fulfil the energy demands and counterbalance risks associated to preterm birth. However, little is known about the potential adverse effects of exposure to persistent organic pollutants (POPs) carried in human milk on preterm infant growth. We conducted a pilot study proving the application of an integrative analytical approach based on mass spectrometry (MS) coupled to advanced statistical models, favouring the comprehensive molecular profiling to support the identification of multiple biomarkers. We applied this workflow in the frame of a preterm infants’ cohort to explore environmental determinants of growth. The combination of high resolution gas and liquid chromatography MS platforms generated a large molecular profile, including 102 pollutants and nutrients (targeted analysis) and 784 metabolites (non-targeted analysis). Data analysis consisted in a preliminary examination of associations between the signatures of POPs and the normalised growth of preterm infants, using multivariate linear regression adjusting for known confounding variables. A second analysis aimed to identify multidimensional biomarkers using a multiblock algorithm allowing the integration of multiple datasets in the growth model of preterm infants. The preliminary results did not suggest an impairment of preterm growth associated to the milk concentrations of POPs. The multiblock approach however revealed complex interrelated molecular networks of POPs, lipids, metabolites and amino acids in breastmilk associated to preterm infant growth, supporting the high potential of biomarkers exploration of this proposed workflow. Whereas the present study intended to identify simultaneously pollutant and nutrient exposure profiles associated to early preterm infant growth, this workflow may be easily adapted and applied to other matrices (e.g. serum) and research settings, favouring the functional exploration of environmental determinants of complex and multifactorial diseases.

Geographical discrimination of red garlic (Allium sativum L.) using fast and non-invasive Attenuated Total Reflectance-Fourier Transformed Infrared (ATR-FTIR) spectroscopy combined with chemometrics

Four varieties of red garlic (Allium sativum L.) cultivated in different Italian territories, Sulmona (Abruzzo), Proceno and Castelliri (Lazio), and Nubia (Sicily), were analysed by Attenuated Total Reflectance-Fourier Transformed Infrared (ATR-FTIR) spectroscopy. ATR-FTIR spectra of bulbils and bulbil tunics were separately acquired and processed by Partial Least Squares Discriminant Analysis (PLS-DA) with the aim of classifying the garlic samples on the basis of their geographical origin. Finally, two multi-block strategies (based on Sequential and Orthogonalized Partial Least Squares and Sequential and Orthogonalized Covariance Selection, coupled with Fisher’s Linear Discriminant Analysis) have been applied in order to test whether a joint analysis of data could lead to higher prediction rates. Eventually, the best results were achieved by the multi-block approach based on SO-PLS, which allows obtaining a total classification rate of 95 % (corresponding to one misclassified sample over 20) in external validation.

Processing of NMR and MS metabolomics data using chemometrics methods: a global tool for fungi biotransformation reactions monitoring.

Biotransformation constitutes an important aspect of the drug discovery process, to mimic human metabolism of active principal ingredient but also to generate new chemical entities. Several microorganisms such as fungi are well adapted to transform drug, whether at the stage of screening or for large-scale production. Due to the high chemical complexity of the biotransformation media, it seems attractive to develop new analytical strategies in order to guarantee an adequate monitoring and optimize the production of targeted metabolites or drug candidates. The model designed for this purpose concerns the biotransformation of a potential histamine H3 antagonist (S38093) in order to produce phase I metabolites. MS, NMR and chemometrics tools were used to monitor biotransformation reactions. First, a screening of eleven filamentous fungi was carried out by UHPLC-UV-MS and principal component analysis to select the best candidates. Subsequently, MS (tR, m/z) and NMR (1H, JRES) fingerprints associated with Consensus OPLS-DA multiblock approach were used to better understand the bioreaction mechanisms in terms of nutrient consumption and hydroxylated metabolites production. Then an experimental design was set up to optimize the production conditions (pH, kinetic) of these target metabolites. This study demonstrates how NMR and MS acquisitions combined with chemometric methods offer an innovative analytical strategy to have a grasp of functionalization mechanisms, and identify metabolites and other compounds (amino acids, nutrients, etc.) in complex biotransformation mixtures.

Chemometric Approach Using ComDim and PLS-DA for Discrimination and Classification of Commercial Yerba Mate (Ilex paraguariensis St. Hil.)

Yerba mate samples from three different states of Brazil were evaluated in order to discriminate them regarding the presence of sugar and geographic origin. High-performance liquid chromatography (HPLC), phytochemical compounds, in vitro antioxidant activity, visible and near-infrared (NIR) spectroscopy, colorimetry, and electronic nose were used in tandem with chemometric methods. The multiblock exploratory analysis (ComDim) was able to discriminate the samples containing sugar; however, it was not possible to discriminate them by geographical origin. Furthermore, ComDim results showed the NIR spectra presented the best discriminating capacity. Partial least square discriminant analysis (PLS-DA) models constructed using NIR spectra classified the samples assertively according to the presence of sugar (100% of sensitivity and specificity for the prediction set), and reasonable models were also obtained for the geographic classification (80% of sensitivity and 93% of specificity for the prediction set). The multiblock approach allowed an overall evaluation of the data collected through different analytical methods. In addition, among the methods applied, NIR spectroscopy was faster and cheaper and allowed for better sample discrimination.

Modelling of low-pressure steam turbines operating at very low flowrates: A multiblock approach

On the basis of experimental and numerical data, physically based relationships have been derived in order to describe the main features of steam flow through the last stages of a multi-stage low pressure turbine. These relationships have been implemented in a multiblock model. Comparison between experimental and calculated data proves that this new tool allows to correctly predict the maximum fluid temperature, the velocity field and the power exchange between fluid and blades when last stages operate at near zero volume flow. This quasi-dimensional approach allows the mechanical and thermo-economic optimization of steam turbine power plants running continuously at low volume flow without using any CFD calculation.

A multi-block sliding approach to calculate the permanent seismic displacement of slopes

The Newmark sliding block method is widely used to assess the seismic performance of earth slopes by calculating the permanent displacement of the simplified sliding mass. The conventional Newmark method models the potential landslide mass as a single rigid block sliding on a well-defined slip surface. In this study, a multi-block approach to calculate the sliding displacement of slopes is presented. The dynamic equation of a series of rigid blocks in different sliding conditions is established and the complex interactions within the slip episodes of rigid block assemblies are explicitly resolved. The results show that the original Newmark single block method may produce unconservative estimates of permanent displacement of the shallow sliding mass when the deep-seated sliding subsequently occurs during the earthquake. The predicted displacement of the upper sliding mass, which is simplified as a summation of the Newmark displacement of each single sliding mass, may either overestimate or underestimate the computed displacement including the coupling of sliding of multiple soil masses. The developed approach extends the well-accepted Newmark sliding block method, and can serve as an alternative to predict the seismic slope displacement in which multiple failure surfaces are expected during earthquakes.

Creating a human head finite element model using a multi-block approach for predicting skull response and brain pressure

To better understand head injuries, human head finite element (FE) models have been reported in the literature. In scenarios where the head is directly impacted and measurements of head accelerations are not available, a high-quality skull model, as well as a high-quality brain model, is needed to predict the effect of impact on the brain through the skull. Furthermore, predicting cranial bone fractures requires comprehensively validated skull models. Lastly, high-quality meshes for both the skull and brain are needed for accurate strain/stress predictions across the entire head. Hence, we adopted a multi-block approach to develop hexahedral meshes for the brain, skull, and scalp simultaneously, a first approach in its kind. We then validated our model against experimental data of brain pressures (Nahum et al., 1977) and comprehensive skull responses (Yoganandan et al., 1995, Yoganandan et al., 2004, and Raymond et al., 2009). We concluded that a human head FE model was developed with capabilities to predict blunt- and ballistic-impact-induced skull fractures and pressure-related brain injuries.

Revisiting Gist-PCA Hashing for Near Duplicate Image Detection

This paper presents a scalable method of near duplicate image detection based on Gist-PCA (principal component analysis) hashing. While most of transform coding methods have been interested in nearest neighbor search with applications to similar image search, we solve a range search problems found in near duplicate detection problems. At first, we argue that the PCA hashing of the Gist descriptor is adequate for near duplicate image detection. Then, we decompose the Gist-PCA binary code into a hash key and a residual binary code for scalability into large-scale datasets. In addition, a multi-block approach is incorporated into the method to deal with strong variations, such as image cropping and border framing. Experimental results show that the proposed method is more accurate and faster than the real-valued Gist descriptor and other nearest neighbor search methods.

Establishing the optimal blocks' order in SO‐PLS: Stepwise SO‐PLS and alternative formulations

AbstractMethods capable of handling multiple blocks of data have increased in popularity both in industry and academia, namely, in the fields of chemistry, food and beverages, pharmacology, and biology. Among the multiblock methodologies, sequential orthogonalized partial least squares (SO‐PLS) has been attracting great interest, given its interpretation capabilities (eg, the possibility to estimate the relative additional contributions of each block to predict the response and the degree of mutual overlap, ie, the blocks' communalities) associated to desirable modeling features, such as the independence from the relative scaling of the different data blocks and the flexibility to handle blocks with different dimensionalities and pseudo‐ranks. Given the sequential nature of SO‐PLS, it is critically dependent on the order of the blocks used. When a priori knowledge exists about the natural ordering of the blocks (eg, data arising from sequential operations in a production process), this specification is straightforward. However, in the absence of such knowledge or in cases where no order should be preferred a priori (as happens in the case study of this article), SO‐PLS faces the problem of having to find the most adequate one through the exhaustive search of all permutations. This problem is particularly relevant when the number of blocks is larger than 3, getting exponentially worse as this number increases. Situations where the number of natural data blocks is significant are already quite frequent (eg in multistage batch operations) and will tend to occur more often in the future, as Manufacturing 4.0 takes its course and data sources from the entire value chain become more abundant. Therefore, more efficient and systematic approaches are required to retain the benefits of SO‐PLS while coping with the increasing demands for data processing and analysis in Big Data applications. In this article, we introduce Stepwise SO‐PLS as an efficient algorithm for selecting the blocks ordering when performing SO‐PLS, with capabilities of block exclusion (a feature not shared by other current multiblock approaches). A robust statistical comparison framework based on Monte Carlo cross‐validation is employed to compare the proposed stepwise SO‐PLS formulation with the current systematic approach for selecting the block order and its new variant allowing for blocks selection, in their prediction capabilities. A real case study of the prediction of Madeira wine age will be used for establishing the comparison.

An a posteriori, efficient, high-spectral resolution hybrid finite-difference method for compressible flows

A high-order hybrid method consisting of a high-accurate explicit finite-difference scheme and a Weighted Essentially Non-Oscillatory (WENO) scheme is proposed in this article. Following this premise, two variants are outlined: a hybrid made up of a Finite Difference scheme and a compact WENO scheme (CRWENO 5), and a hybrid made up of a Finite Difference scheme and a non-compact WENO scheme (WENO 5). The main difference with respect to similar schemes is its a posteriori nature, based on the Multidimensional Optimal Order Detection (MOOD) method. To deal with complex geometries, a multi-block approach using Moving Least Squares (MLS) procedure for communication between meshes is used. The hybrid schemes are validated with several 1D and 2D test cases to illustrate their accuracy and shock-capturing properties.

Relevant and independent multi-block approach for plant-wide process and quality-related monitoring based on KPCA and SVDD

Due to prior knowledge being often unavailable in practice, a multi-block strategy totally based on data-driven analytics is an appropriate alternative for plant-wide processes. However, most recent multi-block methods are relatively vague or insufficient for dividing up the process space and lack the comprehensive fault information for quality-related monitoring. This work intends to develop a more reasonable multi-block method and demonstrate the negative impacts of quality-unrelated variables. Both motivations are entirely dependent on the correlation between variables. A major innovation is to determine those independent or related sets of variables, and to provide a more precise indication for those quality-related faults. Sub-blocks with related variables are each modeled by the KPCA, and the rest of the independent variables are treated as an input for a SVDD model. Finally, all of the statistical indicators are aggregated into a single statistic through Bayesian inference. The benefits of the proposed multi-block scheme (MKPCA-SVDD) are elaborated on in detail using numerical simulation, TE benchmark and industrial p-xylene oxidation process.

МАТЕМАТИЧНЕ МОДЕЛЮВАННЯ ДОЗВУКОВОГО ТУРБУЛЕНТНОГО ОБТІКАННЯ ПРОФІЛЮ NACA 0015, ЩО КОЛИВАЄТЬСЯ

Представлено результати розрахунків профілю NACA 0015, що коливається, для трьох режимів течії: слабкий відрив потоку, що відповідає середньому куту атаки 4°; розвинений відрив – 11° (докритичне обтікання профілю); масивний відрив – 15° (закритичне обтікання профілю). Для чисельного моделювання обтікання профілю, що коливається, застосовуються осереднені за Рейнольдсом нестаціонарні рівняння Нав'є-Стокса нестисливої рідини, замкнуті однопараметричною моделлю турбулентної в'язкості. Розроблений неявний скінчено-об'ємний чисельний алгоритм базується на методі штучної стисливості, що модифікований для розрахунку нестаціонарних задач. Аналізуються отримані поля завихренности, миттєві лінії течії, гістерезисні криві нестаціонарних аеродинамічних навантажень профілю для трьох характерних режимів обтікання. Розроблена методика дозволяє відтворювати структуру нестаціонарного відривного обтікання профілю, що коливається. Відмінності в аеродинамічних характеристиках стаціонарного профілю і профілю, що коливається, при одних і тих же кутах атаки добре узгоджуються з експериментальними даними. Порівняння результатів розрахунків обтікання профілю, що коливається з експериментальними даними і відомими розрахунками інших авторів показало перевагу моделі турбулентності SALSA в порівнянні з іншими моделями, що тестувались особливо при наявності масивного відриву потоку.

Investigation of obstacle effect to improve conjugate heat transfer in backward facing step channel using fast simulation of incompressible flow

The current study investigates the conjugate heat transfer characteristics for laminar flow in backward facing step channel. All of the channel walls are insulated except the lower thick wall under a constant temperature. The upper wall includes a insulated obstacle perpendicular to flow direction. The effect of obstacle height and location on the fluid flow and heat transfer are numerically explored for the Reynolds number in the range of 10 ≤ Re ≤ 300. Incompressible Navier-Stokes and thermal energy equations are solved simultaneously in fluid region by the upwind compact finite difference scheme based on flux-difference splitting in conjunction with artificial compressibility method. In the thick wall, the energy equation is obtained by Laplace equation. A multi-block approach is used to perform parallel computing to reduce the CPU time. Each block is modeled separately by sharing boundary conditions with neighbors. The developed program for modeling was written in FORTRAN language with OpenMP API. The obtained results showed that using of the multi-block parallel computing method is a simple robust scheme with high performance and high-order accurate. Moreover, the obtained results demonstrated that the increment of Reynolds number and obstacle height as well as decrement of horizontal distance between the obstacle and the step improve the heat transfer.

Advanced predictive methods for wine age prediction: Part II – A comparison study of multiblock regression approaches

In this article, we extend the scope of the first paper of the sequel, which was dedicated to the analysis of advanced single-block regression methods (Rendall et al., 2016) [1], to the class of multiblock regression approaches. The datasets contemplated for developing the multiblock approaches are the same as in the former publication: volatile, polyphenols, organic acids composition and the UV–Vis spectra. The context is still the prediction of the ageing time of one of finest Portuguese fortified wines, the Madeira Wine, but now the data collected from the different analytical sources is explored simultaneously, in a more structured and informative way, through multiblock methodologies. The goal of this paper is to provide a critical assessment of a rich variety of multiblock regression methods, namely: Concatenated PLS, Multiblock PLS (MBPLS), Hierarchical PLS (HPLS), Network-Induced Supervised Learning (NI-SL) and Sequential Orthogonalised Partial Least Squares (SO-PLS). A comparison of block scaling methods was also undertaken for the Concatenated PLS algorithm, and new block scaling methods were proposed that led to better prediction performances. This study explores and reveals the potential advantages of applying multiblock methods for fusing datasets from different sources, from both the predictive and interpretability perspectives.

Joint Transceiver Design for Linear MMSE Data Fusion in Coherent MAC Wireless Sensor Networks

This paper considers the design of minimal mean square error (MMSE) transceivers in a wireless sensor network. The problem is nonconvex and challenging, and previous results (with partial solutions and/or with convergence unproved) left much to be desired. Here we propose several approaches – 2 block coordinate descent (BCD), essentially cyclic multi-block method and its variants and distributive method to solve this problem. The proposed 2-BCD approach formulates the subproblem of joint beamformer optimization as a general second-order cone programming problem, which lends itself to standard numerical solvers and which requires no extra assumptions like previous works do. The proposed essentially cyclic multi-block approach further decomposes the joint beamformer design subproblem into multiple blocks, and rigorously solves each with semi-closed-form solution. The distributive algorithm optimizes transmitters in a decentralized manner and has never been considered in existing literature. The distributive algorithm has time complexity independent of number of sensors and is especially suitable for large-scale networks. All the previous BCD-based approaches left some singularity issue unattended as well as the convergence property unaddressed, our proposals are the first to provide a complete and provably-converging analytical solution. Extensive analysis and simulations demonstrate the merits of the novel approaches relative to existing alternatives.

Numerical study of 10-year-old child forearm injury

Forearm fractures in children are very common among all pediatric fractures. However, biomechanical investigations on the pediatric forearm are rather scarce, partially due to the complex anatomy, closely situated, interrelated structures, highly dynamic movement patterns, and lack of appropriate tools. The purpose of this study is to develop a computational tool for child forearm investigation and characterize the mechanical responses of a backward fall using the computational model. A three-dimensional 10-year-old child forearm finite element (FE) model, which includes the ulna, radius, carpal bones, metacarpals, phalanges, cartilages and ligaments, was developed. The high-quality hexahedral FE meshes were created using a multi-block approach to ensure computational accuracy. The material properties of the FE model were obtained by scaling reported adult experimental data. The design of computational experiments was performed to investigate material sensitivity and the effects of relevant parameters in backward fall. Numerical results provided a spectrum of child forearm responses with various effective masses and forearm angles. In addition, a conceptual L-shape wrist guard design was simulated and found to be able to reduce child distal radius fracture.

Lipid characterization of individual porcine oocytes by dual mode DESI-MS and data fusion

The development of sensitive measurements to analyze individual cells is of relevance to elucidate specialized roles or metabolic functions of each cell under physiological and pathological conditions. Lipids play multiple and critical roles in cellular functions and the application of analytical methods in the lipidomics area is of increasing interest. In this work, in vitro maturation of porcine oocytes was studied. Two independent sources of chemical information (represented by mass spectra in the positive and negative ion modes) from single oocytes (immature oocytes, 24-h and 44-h in vitro matured oocytes) were acquired by using desorption electrospray ionization-mass spectrometry (DESI-MS). Low and mid-level data fusion strategies are presented with the aim of better exploring the large amount of chemical information contained in the two mass spectrometric lipid profiles. Data were explored by principal component analysis (PCA) within the two multi-block approaches to include information on free fatty acids, phospholipids, cholesterol-related molecules, di- and triacylglycerols. After data fusion, clearer differences among immature and in vitro matured porcine oocytes were observed, which provide novel information regarding lipid metabolism throughout oocyte maturation. In particular, changes in TAG composition, as well as increase in fatty acid metabolism and membrane complexity were evidenced during the in vitro maturation process. This information can assist the improvement of in vitro embryo production for porcine species.

High-accuracy finite-difference schemes for solving elastodynamic problems in curvilinear coordinates within multiblock approach

We propose highly accurate finite-difference schemes for simulating wave propagation problems described by linear second-order hyperbolic equations. The schemes are based on the summation by parts (SBP) approach modified for applications with violation of input data smoothness. In particular, we derive and implement stable schemes for solving elastodynamic anisotropic problems described by the Navier wave equation in complex geometry. To enhance potential of the method, we use a general type of coordinate transformation and multiblock grids. We also show that the conventional spectral element method (SEM) can be treated as the multiblock finite-difference method whose blocks are the SEM cells with SBP operators on GLL grid.