Modelling the vigour of maize seeds submitted to artificial accelerated ageing based on ATR-FTIR data and chemometric tools (PCA, HCA and PLS-DA)

Deep learning analysis of mid-infrared microscopic imaging data for the diagnosis and classification of human lymphomas

The present study presents an alternative analytical workflow that combines mid-infrared (MIR) microscopic imaging and deep learning to diagnose human lymphoma and differentiate between small and large cell lymphoma. We could show that using a deep learning approach to analyze MIR hyperspectral data obtained from benign and malignant lymph node pathology results in high accuracy for correct classification, learning the distinct region of 3900 to 850 cm-1 . The accuracy is above 95% for every pair of malignant lymphoid tissue and still above 90% for the distinction between benign and malignant lymphoid tissue for binary classification. These results demonstrate that a preliminary diagnosis and subtyping of human lymphoma could be streamlined by applying a deep learning approach to analyze MIR spectroscopic data.

The incorporation of Mg2+ ions into aragonite during biomineralization: Implications for the dolomitization of aragonite

Bacteria can facilitate the increase of Mg²⁺ content in biotic aragonite, but the molecular mechanisms of the incorporation of Mg²⁺ ion into aragonite facilitated by bacteria are still unclear and the dolomitization of aragonite grains is rarely reported. In our laboratory experiments, the content of Mg²⁺ ions in biotic aragonite is higher than that in inorganically-precipitated aragonite and we hypothesize that the higher Mg content may enhance the subsequent dolomitization of aragonite. In this study, biotic aragonite was induced by Bacillus licheniformis Y₁ at different Mg/Ca molar ratios. XRD data show that only aragonite was precipitated in the media with Mg/Ca molar ratios at 6, 9, and 12 after culturing for 25 days. The EDS and atomic absorption results show that the content of Mg²⁺ ions in biotic aragonite increased with rising Mg/Ca molar ratios. In addition, our analyses show that the EPS from the bacteria and the organics extracted from the interior of the biotic aragonite contain the same biomolecules, including Ala, Gly, Glu and hexadecanoic acid. The content of Mg²⁺ ions in the aragonite precipitates mediated by biomolecules is significantly higher than that in inorganically-precipitated aragonite. Additionally, compared with Ala and Gly, the increase of the Mg²⁺ ions content in aragonite promoted by Glu and hexadecanoic acid is more significant. The DFT (density functional theory) calculations reveal that the energy needed for Mg²⁺ ion incorporation into aragonite mediated by Glu, hexadecanoic acid, Gly and Ala increased gradually, but was lower than that without acidic biomolecules. The experiments also show that the Mg²⁺ ion content in the aragonite significantly increased with the increasing concentration of biomolecules. In a medium with high Mg²⁺ concentration and with bacteria, after 2 months, micron-sized dolomite rhombs were precipitated on the surfaces of the aragonite particles. This study may provide new insights into the important role played by biomolecules in the incorporation of the Mg²⁺ ions into aragonite. Moreover, these experiments may contribute towards our understanding of the dolomitization of aragonite in the presence of bacteria.

SPME-GC-MS combined with chemometrics to assess the impact of fermentation time on the components, flavor, and function of Laoxianghuang

Laoxianghuang, fermented from Citrus medica L. var. Sarcodactylis Swingle of the Rutaceae family, is a medicinal food. The volatiles of Laoxianghuang fermented in different years were obtained by solid-phase microextraction combined with gas chromatography–mass spectrometry (SPME-GC–MS). Meanwhile, the evolution of its component-flavor function during the fermentation process was explored in depth by combining chemometrics and performance analyses. To extract the volatile compounds from Laoxianghuang, the fiber coating, extraction time, and desorption temperature were optimized in terms of the number and area of peaks. A polydimethylsiloxane/divinylbenzene (PDMS/DVB) with a thickness of 65 μm fiber, extraction time of 30 min, and desorption temperature of 200 °C were shown to be the optimal conditions. There were 42, 44, 52, 53, 53, and 52 volatiles identified in the 3rd, 5th, 8th, 10th, 15th, and 20th years of fermentation of Laoxianghuang, respectively. The relative contents were 97.87%, 98.50%, 98.77%, 98.85%, 99.08%, and 98.36%, respectively. Terpenes (mainly limonene, γ-terpinene and cymene) displayed the highest relative content and were positively correlated with the year of fermentation, followed by alcohols (mainly α-terpineol, β-terpinenol, and γ-terpineol), ketones (mainly cyclohexanone, D(+)-carvone and β-ionone), aldehydes (2-furaldehyde, 5-methylfurfural, and 1-nonanal), phenols (thymol, chlorothymol, and eugenol), esters (bornyl formate, citronellyl acetate, and neryl acetate), and ethers (n-octyl ether and anethole). Principal component analysis (PCA) and hierarchical cluster analysis (HCA) showed a closer relationship between the composition of Laoxianghuang with similar fermentation years of the same gradient (3rd-5th, 8th-10th, and 15th-20th). Partial least squares discriminant analysis (PLS-DA) VIP scores and PCA-biplot showed that α-terpineol, γ-terpinene, cymene, and limonene were the differential candidate biomarkers. Flavor analysis revealed that Laoxianghuang exhibited wood odor from the 3rd to the 10th year of fermentation, while herb odor appeared in the 15th and the 20th year. This study analyzed the changing pattern of the flavor and function of Laoxianghuang through the evolution of the composition, which provided a theoretical basis for further research on subsequent fermentation.

Evaluation of Green Super Rice Lines for Agronomic and Physiological Traits under Salinity Stress

Rice (Oryza sativa) is an important staple food crop worldwide, especially in east and southeast Asia. About one-third of rice cultivated area is under saline soil, either natural saline soils or irrigation with brackish water. Salinity stress is among the devastating abiotic stresses that not only affect rice growth and crop productivity but also limit its cultivation area globally. Plants adopt multiple tolerance mechanisms at the morphological, physiological, and biochemical levels to tackle salinity stress. To identify these tolerance mechanisms, this study was carried out under both a controlled glass house as well as natural saline field conditions using 22 green super rice (GSR) lines along with two local varieties (“IRRI 6 and Kissan Basmati”). Several morpho-physiological and biochemical parameters along with stress-responsive genes were used as evaluation criteria under normal and salinity stress conditions. Correlation and Principal Component Analysis (PCA) suggested that shoot-related parameters and the salt susceptible index (SSI) can be used for the identification of salt-tolerant genotypes. Based on Agglomerative Hierarchical Cluster (AHC) analysis, two saline-tolerant (“S19 and S20”) and saline-susceptible (“S3 and S24”) lines were selected for further molecular evaluation. Quantitative RT-PCR was performed, and results showed that expression of 1-5-phosphoribosyl -5-5-phosphoribosyl amino methylidene amino imidazole-4-carboxamide isomerase, DNA repair protein recA, and peptide transporter PTR2 related genes were upregulated in salt-tolerant genotypes, suggesting their potential role in salinity tolerance. However, additional validation using reverse genetics approaches will further confirm their specific role in salt tolerance. Identified saline-tolerant lines in this study will be useful genetic resources for future salinity breeding programs.

Longevity of preserved Solanum lycopersicum L. seeds: physicochemical characteristics

Ex situ conservation of plant biodiversity has been increasingly used to prevent further loss of genetic resources. Seed banks, for example, shelter the passport data of germplasm, preserved in detail, and made available for easy access, actions included in the FAO's Second Global Plan. We examined the deterioration of tomato seeds of different varieties stored for 10-year intervals at COMAV's genebank. Samples were analyzed using the conventional Germination and Tetrazolium tests, as well as the non-conventional Differential Scanning Calorimetry and Fourier Transform Infrared Spectrometry techniques, to quickly identify the physiological status of the accessions. Fatty acid profile was also determined. The relationship observed between lipid behavior and seed deterioration under long time storage conditions was the same for both non-conventional and conventional techniques. The viability of the samples was not affected by storage time, however, all the employed methods permitted identifying differences between varieties or accessions of the same variety. The complementary methods helped us interpret a complex data set with many interacting factors, leading to rapid identification of seed quality, increasing processing efficiency in tomato seeds conservation.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s12298-022-01157-9.

PLS-DA and data fusion of visible Reflectance, XRF and FTIR spectroscopy in the classification of mixed historical pigments

In this work samples of historical pigments of green hue were brushed on a canvas and studied by Visible Reflectance, X-Ray Fluorescence and Fourier Transform Infrared Spectroscopy. One hundred samples were investigated, all with green hue, these prepared from pigments themselves green, such as chromium oxide (Cr₂O₃) or from a mixture of pigments that result in green, for example, chrome yellow (PbCrO₄) and Prussian blue (Fe₄[Fe(CN)₆]₃). Because every sample investigated through the spectroscopic techniques were of green hue, the characterization of the pigments present in the mixtures through the visual inspection of spectra has become a complex task in some cases, also, due the large number of recorded spectra. In this work, classification models were developed using the multivariate statistical method Partial Least Squares Discriminant Analysis (PLS-DA) to automate the characterization of the pigments present in the mixtures. The models were developed to classify chromium oxide (Cr₂O₃), chrome yellow (PbCrO₄), cerulean blue (CoO.nSnO₂) and yellow ochre (Fe₂O₃·H₂O + clay + silica). The models were developed from the fusion of data from the three spectroscopic techniques. However, before data fusion, pre-treatments of the spectral data were tested for their influence on the PLS-DA models. The models developed with data from the three techniques made it possible to classify the pigments of interest in the samples with up to 100% effectiveness. The results also indicate that fusion of the data from the three techniques allows to obtain fingerprints of the pigments of interest, which is not always possible using data from only one or two of the techniques applied in this work.

Grade diagnosis of human glioma using Fourier transform infrared microscopy and artificial neural network

The World Health Organization (WHO) grade diagnosis of cancer is essential for surgical outcomes and patient treatment. Traditional pathological grading diagnosis depends on dyes or other histological approaches, and the result interpretation highly relies on the pathologists, making the process time-consuming (>60 min, including the steps of dewaxing to water and H&E staining), resource-wasting, and labor-intensive. In the present study, we report an alternative workflow that combines the Fourier transform infrared (FTIR) microscopy and artificial neural network (ANN) to diagnose the grade of human glioma in a way that is faster (~20 min, including the processes of sample dewaxing, spectra acquisition and analysis), accurate (the prediction accuracy, specificity and sensitivity can reach above 99%), and without reagent. Moreover, this method is much superior to the common classification method of principal component analysis-linear discriminate analysis (PCA-LDA) (the prediction accuracy, specificity and sensitivity are only 87%, 89% and 86%, respectively). The ANN mainly learned the characteristic region of 800-1800 cm^-1 to classify the major histopathologic classes of human glioma. These results demonstrate that the grade diagnosis of human glioma by FTIR microscopy plus ANN can be streamlined, and could serve as a complementary pathway that is independent of the traditional pathology laboratory.

Machine Learning for Seed Quality Classification: An Advanced Approach Using Merger Data from FT-NIR Spectroscopy and X-ray Imaging

Optical sensors combined with machine learning algorithms have led to significant advances in seed science. These advances have facilitated the development of robust approaches, providing decision-making support in the seed industry related to the marketing of seed lots. In this study, a novel approach for seed quality classification is presented. We developed classifier models using Fourier transform near-infrared (FT-NIR) spectroscopy and X-ray imaging techniques to predict seed germination and vigor. A forage grass (Urochloa brizantha) was used as a model species. FT-NIR spectroscopy data and radiographic images were obtained from individual seeds, and the models were created based on the following algorithms: linear discriminant analysis (LDA), partial least squares discriminant analysis (PLS-DA), random forest (RF), naive Bayes (NB), and support vector machine with radial basis (SVM-r) kernel. In the germination prediction, the models individually reached an accuracy of 82% using FT-NIR data, and 90% using X-ray data. For seed vigor, the models achieved 61% and 68% accuracy using FT-NIR and X-ray data, respectively. Combining the FT-NIR and X-ray data, the performance of the classification model reached an accuracy of 85% to predict germination, and 62% for seed vigor. Overall, the models developed using both NIR spectra and X-ray imaging data in machine learning algorithms are efficient in quickly, non-destructively, and accurately identifying the capacity of seed to germinate. The use of X-ray data and the LDA algorithm showed great potential to be used as a viable alternative to assist in the quality classification of U. brizantha seeds.