Search

Near-infrared (NIR) spectroscopy is a proven tool for the determination of food authenticity, mainly because of good classification results and the possibility of industrial use due to its easy and fast application. Since water shows broad absorption bands, the water content of a sample should be as low as possible. Freeze-drying is a commonly used preparatory step for this to reduce the water content in the sample. However, freeze-drying, also known as lyophilization, is very time-consuming impeding the widespread usage of NIR analysis as a rapid method for incoming goods inspections. We used a sample set of 72 almond samples from six economically relevant almond-producing countries to investigate the question of how important lyophilization is to obtain a well-performing classification model. For this approach, the samples were ground and lyophilized for 3 h, 24 h, and 48 h and compared to non-freeze-dried samples. Karl-Fischer titration of non-lyophilized samples showed that water contents ranged from 3.0 to 10.5% and remained constant at 0.36 ± 0.13% after a freeze-drying period of 24 h. The non-freeze-dried samples showed a classification accuracy of 93.9 ± 6.4%, which was in the same range as the samples which were freeze-dried for 3 h (94.2 ± 7.8%), 24 h (92.5 ± 8.7%), and 48 h (95.0 ± 9.0%). Feature selection was performed using the Boruta algorithm, which showed that signals from lipids and proteins are relevant for the origin determination. The presented study showed that samples with low water content, especially nuts, can be analyzed without the time-consuming preparation step of freeze-drying to obtain robust and fast results, which are especially required for incoming goods inspection.

Storage is a critical step in post-harvest processing of hazelnuts since it can lead to mold, rancidity and off-flavor. To comprehensively investigate the effects of hazelnut storage, samples were stored under three short-term and two long-term conditions and the polar metabolome was analyzed by 1H NMR spectroscopy at different time points. Large differences in the polar metabolome were observed between the investigated storage conditions. Classification between samples before and after storage showed that samples stored under harsh conditions for a short-term and under standard conditions for a long-term were almost always classified correctly. Boruta, a feature selection algorithm, was used to identify which metabolites were significant for classification. Increases in hexanoic octanoic and nonanoic acid, all products of lipid oxidation, were found across all conditions and an increase in formic acid in most. Harsh short-term storage resulted in an increase in fumaric and lactic acid, glucose, fructose, and choline and a decrease in acetic acid.

Fourier-transform near-infrared (FT-NIR) spectroscopy was used to determine the geographical origin of 233 hazelnut samples of various varieties from five different countries (Germany, France, Georgia, Italy, Turkey). The experimental determination of the geographical origin of hazelnuts is important, because there are usually large price differences between the producer countries and thus a risk of food fraud that should not be underestimated. The present work is a feasibility study using a low-cost method, as high-field NMR and UPLC-QTOF-MS have already been used for this question. Sample sets were split with repeated nested cross validation and an ensemble of discriminant classifiers with random subspaces was used to build the classification models. By using a preprocessing strategy consisting of multiplicative scatter correction, bucketing and the mean averaging of five measured spectra per sample, a test accuracy of 90.6 ± 3.9% was achieved, which rivals results obtained with much more expensive infrastructure. The application of the feature selection approach surrogate minimal depth showed that the successful classification is mainly caused by protein signals. In addition, a low-level data fusion of the NIR and NMR data was performed to assess how well the two methods complement each other. The data fusion was compared to a complementary approach, where the classification results based on the individual NIR and NMR models were jointly examined. The data fusion performed better than the individual methods with a test accuracy of 96.6 ± 2.8%. A comparison of the outliers in all classification models shows conspicuities in always the same samples, indicating that robust classification models are obtained.

Motivation It has been shown that the machine learning approach random forest can be successfully applied to omics data, such as gene expression data, for classification or regression and to select variables that are important for prediction. However, the complex relationships between predictor variables, in particular between causal predictor variables, make the interpretation of currently applied variable selection techniques difficult. Results Here we propose a new variable selection approach called surrogate minimal depth (SMD) that incorporates surrogate variables into the concept of minimal depth (MD) variable importance. Applying SMD, we show that simulated correlation patterns can be reconstructed and that the increased consideration of variable relationships improves variable selection. When compared with existing state-of-the-art methods and MD, SMD has higher empirical power to identify causal variables while the resulting variable lists are equally stable. In conclusion, SMD is a promising approach to get more insight into the complex interplay of predictor variables and outcome in a high-dimensional data setting. Availability and implementation https://github.com/StephanSeifert/SurrogateMinimalDepth. Supplementary information Supplementary data are available at Bioinformatics online.

Identifying and characterizing the biochemical variation in plant tissues is an important task in many research fields. Small spectral differences of the plant cell wall that are caused by genetic or environmental influences may be superimposed by individual variation as well as by a microscopic heterogeneity in molecular composition and structure of different histological substructures. A set of 56 samples from Cucumis sativus (cucumber) plants, comprising a total of ~168,000 spectra from tissue sections of leaf, stem, and roots was investigated by Raman microspectroscopic mapping excited at 532 nm. A multivariate analysis was carried out in order to assess the variation of the spectra with respect to origin of the tissue, the histological (cell wall) substructures, and the possibility to discriminate the spectra obtained from different individuals that had been subjected to two different conditions during growth. Combining the results of principal component analysis (PCA) based classification with the original spatial information in the maps of 23 sections of leaf xylem, variation in cell wall composition is found for four different individuals that also includes a discrimination of tissue grown in the presence and absence of additional silicic acid in the irrigation water of the plants. The spectral data point to differences in a contribution by carotenoids, as well as by hydroxycinnamic acids to the spectra. The results give new insight into the chemical heterogeneity of plant tissues and may be useful for elucidating biochemical processes associated with biomineralization by vibrational spectroscopy.

Surface-enhanced Raman scattering (SERS) is a valuable analytical technique for the analysis of biological samples. However, due to the nature of SERS it is often challenging to exploit the generated data to obtain the desired information when no reporter or label molecules are used. Here, the suitability of random forest based approaches is evaluated using SERS data generated by a simulation framework that is also presented. More specifically, it is demonstrated that important SERS signals can be identified, the relevance of predefined spectral groups can be evaluated, and the relations of different SERS signals can be analyzed. It is shown that for the selection of important SERS signals Boruta and surrogate minimal depth (SMD) and for the analysis of spectral groups the competing method Learner of Functional Enrichment (LeFE) should be applied. In general, this investigation demonstrates that the combination of random forest approaches and SERS data is very promising for sophisticated analysis of complex biological samples.

Mixtures of pollen grains of three different species (Corylus avellana, Alnus cordata, and Pinus sylvestris) were investigated by matrix-assisted laser desorption/ionization time-of-flight imaging mass spectrometry (MALDI-TOF imaging MS). The amount of pollen grains was reduced stepwise from 10 to single pollen grains. For sample pretreatment, we modified a previously applied approach, where any additional extraction steps were omitted. Our results show that characteristic pollen MALDI mass spectra can be obtained from a single pollen grain, which is the prerequisite for a reliable pollen classification in practical applications. MALDI imaging of laterally resolved pollen grains provides additional information by reducing the complexity of the MS spectra of mixtures, where frequently peak discrimination is observed. Combined with multivariate statistical analyses, such as principal component analysis (PCA), our approach offers the chance for a fast and reliable identification of individual pollen grains by mass spectrometry. Graphical Abstract ᅟ.

Motivation High-throughput technologies allow comprehensive characterization of individuals on many molecular levels. However, training computational models to predict disease status based on omics data is challenging. A promising solution is the integration of external knowledge about structural and functional relationships into the modeling process. We compared four published random forest-based approaches using two simulation studies and nine experimental datasets. Results The self-sufficient prediction error approach should be applied when large numbers of relevant pathways are expected. The competing methods hunting and learner of functional enrichment should be used when low numbers of relevant pathways are expected or the most strongly associated pathways are of interest. The hybrid approach synthetic features is not recommended because of its high false discovery rate. Availability and implementation An R package providing functions for data analysis and simulation is available at GitHub (https://github.com/szymczak-lab/PathwayGuidedRF). An accompanying R data package (https://github.com/szymczak-lab/DataPathwayGuidedRF) stores the processed and quality controlled experimental datasets downloaded from Gene Expression Omnibus (GEO). Supplementary information Supplementary data are available at Bioinformatics online.

Machine learning methods and in particular random forests are promising approaches for prediction based on high dimensional omics data sets. They provide variable importance measures to rank predictors according to their predictive power. If building a prediction model is the main goal of a study, often a minimal set of variables with good prediction performance is selected. However, if the objective is the identification of involved variables to find active networks and pathways, approaches that aim to select all relevant variables should be preferred. We evaluated several variable selection procedures based on simulated data as well as publicly available experimental methylation and gene expression data. Our comparison included the Boruta algorithm, the Vita method, recurrent relative variable importance, a permutation approach and its parametric variant (Altmann) as well as recursive feature elimination (RFE). In our simulation studies, Boruta was the most powerful approach, followed closely by the Vita method. Both approaches demonstrated similar stability in variable selection, while Vita was the most robust approach under a pure null model without any predictor variables related to the outcome. In the analysis of the different experimental data sets, Vita demonstrated slightly better stability in variable selection and was less computationally intensive than Boruta. In conclusion, we recommend the Boruta and Vita approaches for the analysis of high-dimensional data sets. Vita is considerably faster than Boruta and thus more suitable for large data sets, but only Boruta can also be applied in low-dimensional settings.

Summary Fourier transform infrared (FTIR) spectroscopy enables chemical analysis of pollen samples for plant phenotyping to study plant–environment interactions, such as influence of climate change or pathogens. However, current approach, such as microspectroscopy and attenuated total reflection spectroscopy, does not allow for high-throughput protocols. This study at hand suggests a new spectroscopic method for high-throughput characterization of pollen. Samples were measured as thin films of pollen fragments using a Bruker FTIR spectrometer with a high-throughput eXTension (HTS-XT) unit employing 384-well plates. In total, 146 pollen samples, belonging to 31 different pollen species of Fagaceae and Betulaceae and collected during three consecutive years (2012–2014) at locations in Croatia, Germany and Norway, were analysed. Critical steps in the sample preparation and measurement, such as variabilities between technical replicates, between microplates and between spectrometers, were studied. Measurement variations due to sample preparation, microplate holders and instrumentation were low, and thus allowed differentiation of samples with respect to phylogeny and biogeography. The spectral variability for a range of Fagales species (Fagus, Quercus, Betula, Corylus, Alnus and Ostrya) showed high-species-specific differences in pollen's chemical composition due to either location or year. Statistically significant inter-annual and locational differences in the pollen spectra indicate that pollen chemical composition has high phenotypic plasticity and is influenced by local climate conditions. The variations in composition are connected to lipids, proteins, carbohydrates and sporopollenins that play crucial roles in cold and desiccation tolerance, protection against UV radiation and as material and energy reserves. The results of this study demonstrate the value of high-throughput FTIR approach for the systematic collection of data on ecosystems. The novel FTIR approach offers fast, reliable and economical screening of large number of samples by semi-automated methodology. The high-throughput approach could provide crucial understanding on plant–climate interactions with respect to biochemical variation within genera, species and populations.

The analysis of pollen chemical composition is important to many fields, including agriculture, plant physiology, ecology, allergology, and climate studies. Here, the potential of a combination of different spectroscopic and spectrometric methods regarding the characterization of small biochemical differences between pollen samples was evaluated using multivariate statistical approaches. Pollen samples, collected from three populations of the grass Poa alpina, were analyzed using Fourier-transform infrared (FTIR) spectroscopy, Raman spectroscopy, surface enhanced Raman scattering (SERS), and matrix assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS). The variation in the sample set can be described in a hierarchical framework comprising three populations of the same grass species and four different growth conditions of the parent plants for each of the populations. Therefore, the data set can work here as a model system to evaluate the classification and characterization ability of the different spectroscopic and spectrometric methods. ANOVA Simultaneous Component Analysis (ASCA) was applied to achieve a separation of different sources of variance in the complex sample set. Since the chosen methods and sample preparations probe different parts and/or molecular constituents of the pollen grains, complementary information about the chemical composition of the pollen can be obtained. By using consensus principal component analysis (CPCA), data from the different methods are linked together. This enables an investigation of the underlying global information, since complementary chemical data are combined. The molecular information from four spectroscopies was combined with phenotypical information gathered from the parent plants, thereby helping to potentially link pollen chemistry to other biotic and abiotic parameters.

Drugs that influence enzymes of lipid metabolism can cause pathological accumulation of lipids in animal cells. Here, gold nanoparticles, acting as nanosensors that deliver surface-enhanced Raman scattering (SERS) spectra from living cells provide molecular evidence of lipid accumulation in lysosomes after treatment of cultured cells with the three tricyclic antidepressants (TCA) desipramine, amitryptiline, and imipramine. The vibrational spectra elucidate to great detail and with very high sensitivity the composition of the drug-induced lipid accumulations, also observed in fixed samples by electron microscopy and X-ray nanotomography. The nanoprobes show that mostly sphingomyelin is accumulated in the lysosomes but also other lipids, in particular, cholesterol. The observation of sphingomyelin accumulation supports the impairment of the enzyme acid sphingomyelinase. The SERS data were analyzed by random forest based approaches, in particular, by minimal depth variable selection and surrogate minimal depth (SMD), shown here to be particularly useful machine learning tools for the analysis of the lipid signals that contribute only weakly to SERS spectra of cells. SMD is used for the identification of molecular colocalization and interactions of the drug molecules with lipid membranes and for discriminating between the biochemical effects of the three different TCA molecules, in agreement with their different activity. The spectra also indicate that the protein composition is significantly changed in cells treated with the drugs.

The process of silicification in plants and the biochemical effects of silica in plant tissues are largely unknown. To study the molecular changes occurring in growing cells that are exposed to higher than normal concentration of silicic acid, Raman spectra of germinating pollen grains of three species (Pinus nigra, Picea omorika, and Camellia japonica) were analyzed in a multivariate classification approach that takes into account the variation of biochemical composition due to species, plant tissue structure, and germination condition. The results of principal component analyses of the Raman spectra indicate differences in the utilization of stored lipids, a changed mobilization of storage carbohydrates in the pollen grain bodies, and altered composition and/or structure of cellulose of the developing pollen tube cell walls. These biochemical changes vary in the different species.

MALDI time-of-flight mass spectrometry (MALDI-TOF MS) has become a widely used tool for the classification of biological samples. The complex chemical composition of pollen grains leads to highly specific, fingerprint-like mass spectra, with respect to the pollen species. Beyond the species-specific composition, the variances in pollen chemistry can be hierarchically structured, including the level of different populations, of environmental conditions or different genotypes. We demonstrate here the sensitivity of MALDI-TOF MS regarding the adaption of the chemical composition of three Poaceae (grass) pollen for different populations of parent plants by analyzing the mass spectra with partial least squares discriminant analysis (PLS-DA) and principal component analysis (PCA). Thereby, variances in species, population and specific growth conditions of the plants were observed simultaneously. In particular, the chemical pattern revealed by the MALDI spectra enabled discrimination of the different populations of one species. Specifically, the role of environmental changes and their effect on the pollen chemistry of three different grass species is discussed. Analysis of the group formation within the respective populations showed a varying influence of plant genotype on the classification, depending on the species, and permits conclusions regarding the respective rigidity or plasticity towards environmental changes. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) shows adaptation of grass pollen composition

Rationale Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has been suggested as a promising tool for the investigation of pollen, but the usefulness of this approach for classification and identification of pollen species has to be proven by an application to samples of varying taxonomic relations. Method MALDI-MS in combination with hierarchical cluster analysis (HCA) and principal component analysis (PCA) was used to delineate taxonomic relations between plants based on pollen biochemistry. To assess the robustness of the approach, pollen of 74 species of the plant orders Fagales and Coniferales were probed. Results Discrimination at the levels of plant order and genus were achieved using the whole spectral range. In many cases, different species of the same genus could be distinguished. The sources of the spectral/chemical differences at the genus level can be understood using PCA. Specifically, typical mass regions for exact genus detection were identified. Conclusions Our results indicate that the chemical information represented by MALDI-TOFMS data is useful for reconstructing taxonomic relationships and is complementary to other chemical information on pollen from other spectroscopic data. Copyright © 2015 John Wiley & Sons, Ltd.

Matrix-assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF MS) is a well-implemented analytical technique for the investigation of complex biological samples. In MS, the sample preparation strategy is decisive for the success of the measurements. Here, sample preparation processes and target materials for the investigation of different pollen grains are compared. A reduced and optimized sample preparation process prior to MALDI-TOF measurement is presented using conductive carbon tape as target. The application of conductive tape yields in enhanced absolute signal intensities and mass spectral pattern information, which leads to a clear separation in subsequent pattern analysis. The results will be used to improve the taxonomic differentiation and identification, and might be useful for the development of a simple routine method to identify pollen based on mass spectrometry.

By adding a gold core to silica nanoparticles (BrightSilica), silica-like nanoparticles are generated that, unlike unmodified silica nanoparticles, provide three types of complementary information to investigate the silica nano-biointeraction inside eukaryotic cells in situ. Firstly, organic molecules in proximity of and penetrating into the silica shell in live cells are monitored by surface-enhanced Raman scattering (SERS). The SERS data show interaction of the hybrid silica particles with tyrosine, cysteine and phenylalanine side chains of adsorbed proteins. Composition of the biomolecular corona of BrightSilica nanoparticles differs in fibroblast and macrophage cells. Secondly, quantification of the BrightSilica nanoparticles using laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) micromapping indicates a different interaction of silica nanoparticles compared to gold nanoparticles under the same experimental conditions. Thirdly, the metal cores allow the investigation of particle distribution and interaction in the cellular ultrastructure by cryo nanoscale X-ray tomography (cryo-XT). In 3D reconstructions the assumption is confirmed that BrightSilica nanoparticles enter cells by an endocytotic mechanism. The high SERS intensities are explained by the beneficial plasmonic properties due to agglomeration of BrightSilica. The results have implications for the development of multi-modal qualitative and quantitative characterization in comparative nanotoxicology and bionanotechnology.

RATIONALE A fast and reliable online identification of pollen is not yet available. The identification of pollen is based mainly on the evaluation of morphological data obtained by microscopic methods. METHODS Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) was applied to the analysis of extracts and milled pollen samples. The obtained MALDI data were explored for characteristic peak patterns which could be subjected to a multivariate statistical analysis. RESULTS Two sample preparation methods are presented, which require only minimal or no chemical extraction of the pollen. MALDI pollen spectra could be recorded showing various peak patterns. A multivariate statistics approach allowed the classification of pollen into clusters indicating similarities and differences between various species. CONCLUSIONS These results demonstrate the potential and the reliability of MALDI-TOF MS for the identification and, in combination with multivariate statistics, also for the classification of pollen. Copyright © 2012 John Wiley & Sons, Ltd.

Gold and silver nanoparticles can be immobilized on glass slides using aminosilane linkers. Here, we demonstrate that particle monolayer surfaces can also be generated by simultaneous immobilization of both gold and silver nanoparticles with the same organosilane linker. These new surfaces display surface-enhanced Raman scattering (SERS) enhancement typical for gold or silver monolayers, depending on the ratio of the two types of nanoparticles and, at the same time, have the capability to probe complex analytes composed from various molecules which adsorb at only one of the metals. The reported results from scanning electron microscopy, scanning force microscopy, and UV/vis absorbance for surfaces containing one or two types of nanoparticles indicate that an enhancement level above 104 is related to nanoaggregates that form in the 2D plane. High and stable enhancement factors over a wide range of analyte concentrations along with high homogeneity of the enhancement at the microscopic scale make the plasmo...

The molecular orientation and electronic structure of hexa-peri-hexabenzocoronene (HBC) deposited on Ag(111) was investigated in the mono- and multilayer regime by near edge X-ray absorption fine structure spectroscopy (NEXAFS), scanning tunneling and scanning force microscopy (STM and SFM), and ultraviolet photoelectron spectroscopy (UPS). We found that the molecular orientation changed as a function of coverage, exhibiting a transition from flat lying HBC in the monolayer to molecules being 43° inclined relative to the substrate plane in the multilayer. This structural transition was accompanied by a rigid shift of molecular energy levels by 0.3 eV toward higher binding energy between mono- and multilayer. In addition, broadening of the highest molecular levels by ca. 0.35 eV was observed in the multilayer, suggesting the formation of extended intermolecular energy bands.

Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has been suggested as a promising tool for the investigation of pollen, but the usefulness of this approach for classification and identification of pollen species has to be proven by an application to samples of varying taxonomic relations.MALDI-MS in combination with hierarchical cluster analysis (HCA) and principal component analysis (PCA) was used to delineate taxonomic relations between plants based on pollen biochemistry. To assess the robustness of the approach, pollen of 74 species of the plant orders Fagales and Coniferales were probed.Discrimination at the levels of plant order and genus were achieved using the whole spectral range. In many cases, different species of the same genus could be distinguished. The sources of the spectral/chemical differences at the genus level can be understood using PCA. Specifically, typical mass regions for exact genus detection were identified.Our results indicate that the chemical information represented by MALDI-TOFMS data is useful for reconstructing taxonomic relationships and is complementary to other chemical information on pollen from other spectroscopic data.

A fast and reliable online identification of pollen is not yet available. The identification of pollen is based mainly on the evaluation of morphological data obtained by microscopic methods.Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) was applied to the analysis of extracts and milled pollen samples. The obtained MALDI data were explored for characteristic peak patterns which could be subjected to a multivariate statistical analysis.Two sample preparation methods are presented, which require only minimal or no chemical extraction of the pollen. MALDI pollen spectra could be recorded showing various peak patterns. A multivariate statistics approach allowed the classification of pollen into clusters indicating similarities and differences between various species.These results demonstrate the potential and the reliability of MALDI-TOF MS for the identification and, in combination with multivariate statistics, also for the classification of pollen.