Artificial neural networks for classification in metabolomic studies of whole cells using 1H nuclear magnetic resonance

Prediction of multiple pH compartments by deep learning in magnetic resonance spectroscopy with hyperpolarized 13C-labelled zymonic acid

BACKGROUND

Hyperpolarization enhances the sensitivity of nuclear magnetic resonance experiments by between four and five orders of magnitude. Several hyperpolarized sensor molecules have been introduced that enable high sensitivity detection of metabolism and physiological parameters. However, hyperpolarized magnetic resonance spectroscopy imaging (MRSI) often suffers from poor signal-to-noise ratio and spectral analysis is complicated by peak overlap. Here, we study measurements of extracellular pH (pHe) by hyperpolarized zymonic acid, where multiple pHe compartments, such as those observed in healthy kidney or other heterogeneous tissue, result in a cluster of spectrally overlapping peaks, which is hard to resolve with conventional spectroscopy analysis routines.

METHODS

We investigate whether deep learning methods can yield improved pHe prediction in hyperpolarized zymonic acid spectra of multiple pHe compartments compared to conventional line fitting. As hyperpolarized 13C-MRSI data sets are often small, a convolutional neural network (CNN) and a multilayer perceptron (MLP) were trained with either a synthetic or a mixed (synthetic and augmented) data set of acquisitions from the kidneys of healthy mice.

RESULTS

Comparing the networks' performances compartment-wise on a synthetic test data set and eight real kidney data shows superior performance of CNN compared to MLP and equal or superior performance compared to conventional line fitting. For correct prediction of real kidney pHe values, training with a mixed data set containing only 0.5% real data shows a large improvement compared to training with synthetic data only. Using a manual segmentation approach, pH maps of kidney compartments can be improved by neural network predictions for voxels including three pH compartments.

CONCLUSION

The results of this study indicate that CNNs offer a reliable, accurate, fast and non-interactive method for analysis of hyperpolarized 13C MRS and MRSI data, where low amounts of acquired data can be complemented to achieve suitable network training.

NMR in Metabolomics: From Conventional Statistics to Machine Learning and Neural Network Approaches

NMR measurements combined with chemometrics allow achieving a great amount of information for the identification of potential biomarkers responsible for a precise metabolic pathway. These kinds of data are useful in different fields, ranging from food to biomedical fields, including health science. The investigation of the whole set of metabolites in a sample, representing its fingerprint in the considered condition, is known as metabolomics and may take advantage of different statistical tools. The new frontier is to adopt self-learning techniques to enhance clustering or classification actions that can improve the predictive power over large amounts of data. Although machine learning is already employed in metabolomics, deep learning and artificial neural networks approaches were only recently successfully applied. In this work, we give an overview of the statistical approaches underlying the wide range of opportunities that machine learning and neural networks allow to perform with accurate metabolites assignment and quantification.Various actual challenges are discussed, such as proper metabolomics, deep learning architectures and model accuracy.

Study on toxicity effects of environmental pollutants based on metabolomics: A review

In the past few decades, the toxic effects of environmental pollutants on non-target organisms have received more and more attention. As a new omics technology, metabolomics can clarify the metabolic homeostasis of the organism at the overall level by studying the changes in the relative contents of endogenous metabolites in the organism. Recently, a large number of studies have used metabolomics technology to study the toxic effects of environmental pollutants on organisms. In this review, we reviewed the analysis processes and data processes of metabolomics and its application in the study of the toxic effects of environmental pollutants including heavy metals, pesticides, polychlorinated biphenyls, polycyclic aromatic hydrocarbons, polybrominated diphenyl ethers and microplastics. In addition, we emphasized that the combination of metabolomics and other omics technologies will help to explore the toxic mechanism of environmental pollutants and provide new research ideas for the toxicological evaluation of environmental pollutants.

Identification of flux checkpoints in a metabolic pathway through white-box, grey-box and black-box modeling approaches

Metabolic pathway modeling plays an increasing role in drug design by allowing better understanding of the underlying regulation and controlling networks in the metabolism of living organisms. However, despite rapid progress in this area, pathway modeling can become a real nightmare for researchers, notably when few experimental data are available or when the pathway is highly complex. Here, three different approaches were developed to model the second part of glycolysis of E. histolytica as an application example, and have succeeded in predicting the final pathway flux: one including detailed kinetic information (white-box), another with an added adjustment term (grey-box) and the last one using an artificial neural network method (black-box). Afterwards, each model was used for metabolic control analysis and flux control coefficient determination. The first two enzymes of this pathway are identified as the key enzymes playing a role in flux control. This study revealed the significance of the three methods for building suitable models adjusted to the available data in the field of metabolic pathway modeling, and could be useful to biologists and modelers.

Computational Methods for the Discovery of Metabolic Markers of Complex Traits

Metabolomics uses quantitative analyses of metabolites from tissues or bodily fluids to acquire a functional readout of the physiological state. Complex diseases arise from the influence of multiple factors, such as genetics, environment and lifestyle. Since genes, RNAs and proteins converge onto the terminal downstream metabolome, metabolomics datasets offer a rich source of information in a complex and convoluted presentation. Thus, powerful computational methods capable of deciphering the effects of many upstream influences have become increasingly necessary. In this review, the workflow of metabolic marker discovery is outlined from metabolite extraction to model interpretation and validation. Additionally, current metabolomics research in various complex disease areas is examined to identify gaps and trends in the use of several statistical and computational algorithms. Then, we highlight and discuss three advanced machine-learning algorithms, specifically ensemble learning, artificial neural networks, and genetic programming, that are currently less visible, but are budding with high potential for utility in metabolomics research. With an upward trend in the use of highly-accurate, multivariate models in the metabolomics literature, diagnostic biomarker panels of complex diseases are more recently achieving accuracies approaching or exceeding traditional diagnostic procedures. This review aims to provide an overview of computational methods in metabolomics and promote the use of up-to-date machine-learning and computational methods by metabolomics researchers.

eDoctor: machine learning and the future of medicine

Machine learning (ML) is a burgeoning field of medicine with huge resources being applied to fuse computer science and statistics to medical problems. Proponents of ML extol its ability to deal with large, complex and disparate data, often found within medicine and feel that ML is the future for biomedical research, personalized medicine, computer-aided diagnosis to significantly advance global health care. However, the concepts of ML are unfamiliar to many medical professionals and there is untapped potential in the use of ML as a research tool. In this article, we provide an overview of the theory behind ML, explore the common ML algorithms used in medicine including their pitfalls and discuss the potential future of ML in medicine.

Influence of media selection on NMR based metabolic profiling of human cell lines

INTRODUCTION

Comparative metabolic profiling of different human cancer cell lines can reveal metabolic pathways up-regulated or down-regulated in each cell line, potentially providing insight into distinct metabolism taking place in different types of cancer cells. It is noteworthy, however, that human cell lines available from public repositories are deposited with recommended media for optimal growth, and if cell lines to be compared are cultured on different growth media, this introduces a potentially serious confounding variable in metabolic profiling studies designed to identify intrinsic metabolic pathways active in each cell line.

OBJECTIVES

The goal of this study was to determine if the culture media used to grow human cell lines had a significant impact on the measured metabolic profiles.

METHODS

NMR-based metabolic profiles of hydrophilic extracts of three human pancreatic cancer cell lines, AsPC-1, MiaPaCa-2 and Panc-1, were compared after culture on Dulbecco's Modified Eagle Medium (DMEM) or Roswell Park Memorial Institute (RPMI-1640) medium.

RESULTS

Comparisons of the same cell lines cultured on different media revealed that the concentrations of many metabolites depended strongly on the choice of culture media. Analyses of different cell lines grown on the same media revealed insight into their metabolic differences.

CONCLUSION

The choice of culture media can significantly impact metabolic profiles of human cell lines and should be considered an important variable when designing metabolic profiling studies. Also, the metabolic differences of cells cultured on media recommended for optimal growth in comparison to a second growth medium can reveal critical insight into metabolic pathways active in each cell line.

Machine Learning Methods for Analysis of Metabolic Data and Metabolic Pathway Modeling

Machine learning uses experimental data to optimize clustering or classification of samples or features, or to develop, augment or verify models that can be used to predict behavior or properties of systems. It is expected that machine learning will help provide actionable knowledge from a variety of big data including metabolomics data, as well as results of metabolism models. A variety of machine learning methods has been applied in bioinformatics and metabolism analyses including self-organizing maps, support vector machines, the kernel machine, Bayesian networks or fuzzy logic. To a lesser extent, machine learning has also been utilized to take advantage of the increasing availability of genomics and metabolomics data for the optimization of metabolic network models and their analysis. In this context, machine learning has aided the development of metabolic networks, the calculation of parameters for stoichiometric and kinetic models, as well as the analysis of major features in the model for the optimal application of bioreactors. Examples of this very interesting, albeit highly complex, application of machine learning for metabolism modeling will be the primary focus of this review presenting several different types of applications for model optimization, parameter determination or system analysis using models, as well as the utilization of several different types of machine learning technologies.

Prediction of Clinical Endpoints in Breast Cancer Using NMR Metabolic Profiles

Metabolic profiles reflect biological conditions as a result of biochemical changes within a living system. It is therefore possible to associate metabolic signatures with clinical endpoints of diseases, such as breast cancer. Nuclear magnetic resonance (NMR) spectroscopy is one of the most common techniques used for metabolic profiling, and produces high dimensional datasets from which meaningful biological information can be extracted. Here, we present an overview of data analysis techniques used to achieve this, describing key steps in the procedure. Moreover, examples of clinical endpoints of interest are provided. Although these are specific for breast cancer, the procedures for the analysis of NMR spectra as described here are applicable to any type of cancer and to other diseases.

Integrative metabolomics as emerging tool to study autophagy regulation

Recent technological developments in metabolomics research have enabled in-depth characterization of complex metabolite mixtures in a wide range of biological, biomedical, environmental, agricultural, and nutritional research fields. Nuclear magnetic resonance spectroscopy and mass spectrometry are the two main platforms for performing metabolomics studies. Given their broad applicability and the systemic insight into metabolism that can be obtained it is not surprising that metabolomics becomes increasingly popular in basic biological research. In this review, we provide an overview on key metabolites, recent studies, and future opportunities for metabolomics in studying autophagy regulation. Metabolites play a pivotal role in autophagy regulation and are therefore key targets for autophagy research. Given the recent success of metabolomics, it can be expected that metabolomics approaches will contribute significantly to deciphering the complex regulatory mechanisms involved in autophagy in the near future and promote understanding of autophagy and autophagy-related diseases in living cells and organisms.

Evaluation of Classifier Performance for Multiclass Phenotype Discrimination in Untargeted Metabolomics

Statistical classification is a critical component of utilizing metabolomics data for examining the molecular determinants of phenotypes. Despite this, a comprehensive and rigorous evaluation of the accuracy of classification techniques for phenotype discrimination given metabolomics data has not been conducted. We conducted such an evaluation using both simulated and real metabolomics datasets, comparing Partial Least Squares-Discriminant Analysis (PLS-DA), Sparse PLS-DA, Random Forests, Support Vector Machines (SVM), Artificial Neural Network, k-Nearest Neighbors (k-NN), and Naïve Bayes classification techniques for discrimination. We evaluated the techniques on simulated data generated to mimic global untargeted metabolomics data by incorporating realistic block-wise correlation and partial correlation structures for mimicking the correlations and metabolite clustering generated by biological processes. Over the simulation studies, covariance structures, means, and effect sizes were stochastically varied to provide consistent estimates of classifier performance over a wide range of possible scenarios. The effects of the presence of non-normal error distributions, the introduction of biological and technical outliers, unbalanced phenotype allocation, missing values due to abundances below a limit of detection, and the effect of prior-significance filtering (dimension reduction) were evaluated via simulation. In each simulation, classifier parameters, such as the number of hidden nodes in a Neural Network, were optimized by cross-validation to minimize the probability of detecting spurious results due to poorly tuned classifiers. Classifier performance was then evaluated using real metabolomics datasets of varying sample medium, sample size, and experimental design. We report that in the most realistic simulation studies that incorporated non-normal error distributions, unbalanced phenotype allocation, outliers, missing values, and dimension reduction, classifier performance (least to greatest error) was ranked as follows: SVM, Random Forest, Naïve Bayes, sPLS-DA, Neural Networks, PLS-DA and k-NN classifiers. When non-normal error distributions were introduced, the performance of PLS-DA and k-NN classifiers deteriorated further relative to the remaining techniques. Over the real datasets, a trend of better performance of SVM and Random Forest classifier performance was observed.

Multivariate Calibration Approach for Quantitative Determination of Cell-Line Cross Contamination by Intact Cell Mass Spectrometry and Artificial Neural Networks

Cross-contamination of eukaryotic cell lines used in biomedical research represents a highly relevant problem. Analysis of repetitive DNA sequences, such as Short Tandem Repeats (STR), or Simple Sequence Repeats (SSR), is a widely accepted, simple, and commercially available technique to authenticate cell lines. However, it provides only qualitative information that depends on the extent of reference databases for interpretation. In this work, we developed and validated a rapid and routinely applicable method for evaluation of cell culture cross-contamination levels based on mass spectrometric fingerprints of intact mammalian cells coupled with artificial neural networks (ANNs). We used human embryonic stem cells (hESCs) contaminated by either mouse embryonic stem cells (mESCs) or mouse embryonic fibroblasts (MEFs) as a model. We determined the contamination level using a mass spectra database of known calibration mixtures that served as training input for an ANN. The ANN was then capable of correct quantification of the level of contamination of hESCs by mESCs or MEFs. We demonstrate that MS analysis, when linked to proper mathematical instruments, is a tangible tool for unraveling and quantifying heterogeneity in cell cultures. The analysis is applicable in routine scenarios for cell authentication and/or cell phenotyping in general.

Automatic NMR-based identification of chemical reaction types in mixtures of co-occurring reactions

The combination of chemoinformatics approaches with NMR techniques and the increasing availability of data allow the resolution of problems far beyond the original application of NMR in structure elucidation/verification. The diversity of applications can range from process monitoring, metabolic profiling, authentication of products, to quality control. An application related to the automatic analysis of complex mixtures concerns mixtures of chemical reactions. We encoded mixtures of chemical reactions with the difference between the ¹H NMR spectra of the products and the reactants. All the signals arising from all the reactants of the co-occurring reactions were taken together (a simulated spectrum of the mixture of reactants) and the same was done for products. The difference spectrum is taken as the representation of the mixture of chemical reactions. A data set of 181 chemical reactions was used, each reaction manually assigned to one of 6 types. From this dataset, we simulated mixtures where two reactions of different types would occur simultaneously. Automatic learning methods were trained to classify the reactions occurring in a mixture from the ¹H NMR-based descriptor of the mixture. Unsupervised learning methods (self-organizing maps) produced a reasonable clustering of the mixtures by reaction type, and allowed the correct classification of 80% and 63% of the mixtures in two independent test sets of different similarity to the training set. With random forests (RF), the percentage of correct classifications was increased to 99% and 80% for the same test sets. The RF probability associated to the predictions yielded a robust indication of their reliability. This study demonstrates the possibility of applying machine learning methods to automatically identify types of co-occurring chemical reactions from NMR data. Using no explicit structural information about the reactions participants, reaction elucidation is performed without structure elucidation of the molecules in the mixtures.

Hyperspectral imaging based method for fast characterization of kidney stone types

The formation of kidney stones is a common and highly studied disease, which causes intense pain and presents a high recidivism. In order to find the causes of this problem, the characterization of the main compounds is of great importance. In this sense, the analysis of the composition and structure of the stone can give key information about the urine parameters during the crystal growth. But the usual methods employed are slow, analyst dependent and the information obtained is poor. In the present work, the near infrared (NIR)-hyperspectral imaging technique was used for the analysis of 215 samples of kidney stones, including the main types usually found and their mixtures. The NIR reflectance spectra of the analyzed stones showed significant differences that were used for their classification. To do so, a method was created by the use of artificial neural networks, which showed a probability higher than 90% for right classification of the stones. The promising results, robust methodology, and the fast analytical process, without the need of an expert assistance, lead to an easy implementation at the clinical laboratories, offering the urologist a rapid diagnosis that shall contribute to minimize urolithiasis recidivism.

Bioinformatics Tools for Mass Spectroscopy-Based Metabolomic Data Processing and Analysis

Biological systems are increasingly being studied in a holistic manner, using omics approaches, to provide quantitative and qualitative descriptions of the diverse collection of cellular components. Among the omics approaches, metabolomics, which deals with the quantitative global profiling of small molecules or metabolites, is being used extensively to explore the dynamic response of living systems, such as organelles, cells, tissues, organs and whole organisms, under diverse physiological and pathological conditions. This technology is now used routinely in a number of applications, including basic and clinical research, agriculture, microbiology, food science, nutrition, pharmaceutical research, environmental science and the development of biofuels. Of the multiple analytical platforms available to perform such analyses, nuclear magnetic resonance and mass spectrometry have come to dominate, owing to the high resolution and large datasets that can be generated with these techniques. The large multidimensional datasets that result from such studies must be processed and analyzed to render this data meaningful. Thus, bioinformatics tools are essential for the efficient processing of huge datasets, the characterization of the detected signals, and to align multiple datasets and their features. This paper provides a state-of-the-art overview of the data processing tools available, and reviews a collection of recent reports on the topic. Data conversion, pre-processing, alignment, normalization and statistical analysis are introduced, with their advantages and disadvantages, and comparisons are made to guide the reader.

Metabolomics in the study of kidney diseases

Metabolomics--the nontargeted measurement of all metabolites produced by the body--is beginning to show promise in both biomarker discovery and, in the form of pharmacometabolomics, in aiding the choice of therapy for patients with specific diseases. In its two basic forms (pattern recognition and metabolite identification), this developing field has been used to discover potential biomarkers in several renal diseases, including acute kidney injury (attributable to a variety of causes), autosomal dominant polycystic kidney disease and kidney cancer. NMR and gas chromatography or liquid chromatography, together with mass spectrometry, are generally used to separate and identify metabolites. Many hurdles need to be overcome in this field, such as achieving consistency in collection of biofluid samples, controlling for batch effects during the analysis and applying the most appropriate statistical analysis to extract the maximum amount of biological information from the data obtained. Pathway and network analyses have both been applied to metabolomic analysis, which vastly extends its clinical relevance and effects. In addition, pharmacometabolomics analyses, in which a metabolomic signature can be associated with a given therapeutic effect, are beginning to appear in the literature, which will lead to personalized therapies. Thus, metabolomics holds promise for early diagnosis, increased choice of therapy and the identification of new metabolic pathways that could potentially be targeted in kidney disease.

Metabolomics in the study of kidney diseases

Metabolomics--the nontargeted measurement of all metabolites produced by the body--is beginning to show promise in both biomarker discovery and, in the form of pharmacometabolomics, in aiding the choice of therapy for patients with specific diseases. In its two basic forms (pattern recognition and metabolite identification), this developing field has been used to discover potential biomarkers in several renal diseases, including acute kidney injury (attributable to a variety of causes), autosomal dominant polycystic kidney disease and kidney cancer. NMR and gas chromatography or liquid chromatography, together with mass spectrometry, are generally used to separate and identify metabolites. Many hurdles need to be overcome in this field, such as achieving consistency in collection of biofluid samples, controlling for batch effects during the analysis and applying the most appropriate statistical analysis to extract the maximum amount of biological information from the data obtained. Pathway and network analyses have both been applied to metabolomic analysis, which vastly extends its clinical relevance and effects. In addition, pharmacometabolomics analyses, in which a metabolomic signature can be associated with a given therapeutic effect, are beginning to appear in the literature, which will lead to personalized therapies. Thus, metabolomics holds promise for early diagnosis, increased choice of therapy and the identification of new metabolic pathways that could potentially be targeted in kidney disease.