Dataset Information

Deep reinforcement learning classification of sparkling wines based on ICP-MS and DOSY NMR spectra.

ABSTRACT: An approach that combines NMR spectroscopy and inductively coupled plasma mass spectrometry (ICP-MS) and advanced tensor decomposition algorithms with state-of-the-art deep learning procedures was applied for the classification of Croatian continental sparkling wines by their geographical origin. It has been demonstrated that complex high-dimensional NMR or ICP-MS data cannot be classified by higher-order tensor decomposition alone. Extension of the procedure by deep reinforcement learning resulted in an exquisite neural network predictive model for the classification of sparkling wines according to their geographical origin. A network trained on half of the sample set was able to classify even 94% of all samples. The model can particularly be useful in cases where the number of samples is limited and when simpler statistical methods fail to produce reliable data. The model can further be exploited for the identification and differentiation of sparkling wines including a high potential for authenticity or quality control.

SUBMITTER: Jagatic Korenika AM

PROVIDER: S-EPMC10847605 | biostudies-literature | 2024 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep reinforcement learning classification of sparkling wines based on ICP-MS and DOSY NMR spectra.

Jagatić Korenika Ana-Marija AM Jeromel Ana A Tomaz Ivana I Jednačak Tomislav T Rončević Sanda S Nemet Ivan I Primožič Ines I Hrenar Tomica T Novak Predrag P

Food chemistry: X 20240128

An approach that combines NMR spectroscopy and inductively coupled plasma mass spectrometry (ICP-MS) and advanced tensor decomposition algorithms with <i>state-of-the-art</i> deep learning procedures was applied for the classification of Croatian continental sparkling wines by their geographical origin. It has been demonstrated that complex high-dimensional NMR or ICP-MS data cannot be classified by higher-order tensor decomposition alone. Extension of the procedure by deep reinforcement learnin ...[more]

PMID: 38328694

Similar Datasets

Project description:Sparkling wines elaboration has been studied by several research groups, but this is the first report on analysis of biological processes according to the Gene Ontology terms (GO terms) and related to proteins expressed by yeast cells during the second fermentation of sparkling wines. This work provides a comprehensive study of the most relevant biological processes in Saccharomyces cerevisiae P29, a sparkling wine strain, during the second fermentation under two conditions (without and with endogenous CO2 overpressure) in the middle and the end of second fermentation. Consequently, a proteomic analysis with the OFFGEL fractionator and protein identification with LTQ Orbitrap XL coupled to HPLC were performed. The classification of biological processes was carried out using the tools provided by the Saccharomyces Genome Database. Results indicate that a greater number of biological processes were identified under condition without CO2 overpressure and in the middle of the fermentation versus the end of the second fermentation. The biological processes highlighted under condition without CO2 overpressure in the middle of the fermentation were involved in the carbohydrate and lipid metabolic processes and catabolic and biosynthetic processes. However, under CO2 overpressure, specific protein expression in response to stress, transport, translation, and chromosome organization and specific processes were not found. At the end of fermentation, there were higher specific processes under condition without CO2 overpressure; most were related to cell division, growth, biosynthetic process, and gene transcription resulting in increased cell viability in this condition. Under CO2 overpressure condition, the most representative processes were related to translation as tRNA metabolic process, chromosome organization, mRNA processing, ribosome biogenesis, and ribonucleoprotein complex assembly, probably in response to the stress caused by the hard fermentation conditions. Therefore, a broader knowledge of the adaptation of the yeast, and its behavior under typical conditions to produce sparkling wine, might improve and favor the wine industry and the selection of yeast for obtaining a high-quality wine.

Project description:The objective of this research focuses on the development of a statistical methodology able to answer the question of whether variation in the intake of sulfur amino acids (SAA) affects the metabolic process. Traditional approaches, which evaluate specific biomarkers after a series of preprocessing procedures, have been criticized as not being fully informative, as well as inappropriate for translation of methodology. Rather than focusing on particular biomarkers, our proposed methodology involves the multifractal analysis that measures the inhomogeneity of regularity of the proton nuclear magnetic resonance (1H-NMR) spectrum by wavelet-based multifractal spectrum. With two different statistical models (Model-I and Model-II), three different geometric features of the multifractal spectrum of each 1H-NMR spectrum (spectral mode, left slope, and broadness) are employed to evaluate the effect of SAA and discriminate 1H-NMR spectra associated with different treatments. The investigated effects of SAA include group effect (high and low doses of SAA), depletion/repletion effect, and time over data effect. The 1H-NMR spectra analysis outcomes show that group effect is significant for both models. The hourly variation in time and depletion/repletion effects does not show noticeable differences for the three features in Model-I. However, these two effects are significant for the spectral mode feature in Model-II. The 1H-NMR spectra of the SAA low groups exhibit highly regular patterns with more variability than that of the SAA high groups for both models. Moreover, the discriminatory analysis conducted using the support vector machine and the principal components analysis shows that the 1H-NMR spectra of SAA high and low groups can be easily discriminatory for both models, while the spectra of depletion and repletion within these groups are discriminatory for Model-I and Model-II. Therefore, the study outcomes imply that the amount of SAA is important and that SAA intake affects mostly the hourly variation of the metabolic process and the difference between depletion and repletion each day. In conclusion, the proposed multifractal analysis of 1H-NMR spectra provides a novel tool to investigate metabolic processes.

Project description:Diffusion-ordered NMR spectroscopy (DOSY) can be used to analyze mixtures of compounds since resonances deriving from different compounds are distinguished by their diffusion coefficients (D). Previously, DOSY has mostly been used for organometallic and polymer analysis, we have now applied DOSY to investigate diffusion coefficients of structurally diverse organic compounds such as natural products (NP). The experimental Ds derived from 55 diverse NPs has allowed us to establish a power law relationship between D and molecular weight (MW) and therefore predict MW from experimental D. We have shown that D is also affected by factors such as hydrogen bonding, molar density and molecular shape of the compound and we have generated new models that incorporate experimentally derived variables for these factors so that more accurate predictions of MW can be calculated from experimental D. The recognition that multiple physicochemical properties affect D has allowed us to generate a polynomial equation based on multiple linear regression analysis of eight calculated physicochemical properties from 63 compounds to accurately correlate predicted D with experimental D for any known organic compound. This equation has been used to calculate predicted D for 217 043 compounds present in a publicly available natural product database (DEREP-NP) and to dereplicate known NPs in a mixture based on matching of experimental D and structural features derived from NMR analysis with predicted D and calculated structural features in the database. These models have been validated by the dereplication of a mixture of two known sesquiterpenes obtained from Tasmannia xerophila and the identification of new alkaloids from the bryozoan Amathia lamourouxi. These new methodologies allow the MW of compounds in mixtures to be predicted without the need for MS analysis, the dereplication of known compounds and identification of new compounds based solely on parameters derived by DOSY NMR.

Dataset Information

Deep reinforcement learning classification of sparkling wines based on ICP-MS and DOSY NMR spectra.

Publications

Deep reinforcement learning classification of sparkling wines based on ICP-MS and DOSY NMR spectra.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets