Deconvolution of 1D NMR spectra: A deep learning-based approach

A Machine Learning Approach for Efficiently Predicting Polymer Aging from UV-Vis Spectra

This research has introduced an innovative approach that proficiently forecasts the alterations in ultraviolet-visible spectroscopy (UV-Vis) of polymer solutions during the aging effect. This method combines readily accessible feature descriptors with classical machine learning (ML) algorithms. Traditional spectral measurements, while precise in analyzing physical properties, are limited by their cost and efficiency. Therefore, this paper introduces a method that utilizes wavelength and the blue (B), green (G), and red (R) color values of the solutions as input features. We employed seven different ML models to train on these features with 10-fold cross-validation to ensure the reliability and generalizability of our results. After comparative analysis, all of the models performed excellently. Among them, the ExtraTree model demonstrated particularly high precision and excellent predictive ability on the testing set, with a Pearson correlation coefficient (r) of 0.9859 and a mean absolute error (MAE) of 0.0457. This study offers a practical solution for the rapid and cost-effective evaluation of polymer solutions' aging effect.

SPINNED: Simulation-based physics-informed neural network for deconvolution of dynamic susceptibility contrast MRI perfusion data

PURPOSE

To propose the simulation-based physics-informed neural network for deconvolution of dynamic susceptibility contrast (DSC) MRI (SPINNED) as an alternative for more robust and accurate deconvolution compared to existing methods.

METHODS

The SPINNED method was developed by generating synthetic tissue residue functions and arterial input functions through mathematical simulations and by using them to create synthetic DSC MRI time series. The SPINNED model was trained using these simulated data to learn the underlying physical relation (deconvolution) between the DSC-MRI time series and the arterial input functions. The accuracy and robustness of the proposed SPINNED method were assessed by comparing it with two common deconvolution methods in DSC MRI data analysis, circulant singular value decomposition, and Volterra singular value decomposition, using both simulation data and real patient data.

RESULTS

The proposed SPINNED method was more accurate than the conventional methods across all SNR levels and showed better robustness against noise in both simulation and real patient data. The SPINNED method also showed much faster processing speed than the conventional methods.

CONCLUSION

These results support that the proposed SPINNED method can be a good alternative to the existing methods for resolving the deconvolution problem in DSC MRI. The proposed method does not require any separate ground-truth measurement for training and offers additional benefits of quick processing time and coverage of diverse clinical scenarios. Consequently, it will contribute to more reliable, accurate, and rapid diagnoses in clinical applications compared with the previous methods including those based on supervised learning.

Application of Odd-Order Derivatives in Fourier Transform Nuclear Magnetic Resonance Spectroscopy toward Quantitative Deconvolution

When Fourier transform (FT) spectrum peaks are overlapped, primary maxima of odd-order derivatives can be used to evaluate their independent intensities. We studied the feasibility of higher odd-order derivatives on Lorentzian peak shape and magnitude peak shape. Simulation studies for FT nuclear magnetic resonance (NMR) spectroscopy demonstrated good results toward quantitative deconvolution of overlapping FT spectrum peaks. Although it is not so desirable to deconvolute special line shapes such as Gaussian, Voigt, and Tsallis profiles, the odd-order derivatives exhibit a bright future compared to even-order derivatives. An application example of practical NMR spectroscopy with ethylbenzene isomers is presented. White Gaussian noises were added to the simulated spectra at two different signal-to-noise ratios (20 and 40). Kauppinen's denoising and smoothing algorithms can effectively remove interference of the noise and help to have good deconvoluting results using the odd-order derivatives. We compared features of our approach with popular deconvolution sharpening algorithms and conducted a comparison study with Kauppinen's Fourier self-deconvolution. Our approach has a better dynamic range of peak intensities and is not sensitive to the sampling rates. Other common deconvolution methods are also discussed briefly.

Bayesian analysis of 1D 1H-NMR spectra

Extracting spin system parameters from 1D high resolution 1H-NMR spectra can be an intricate task requiring sophisticate methods. With a few exceptions methods to perform such a total line shape analysis commonly rely on local optimization techniques which for increasing complexity of the underlying spin system tend to reveal local solutions. In this work we propose a full Bayesian modeling approach based on a quantum mechanical model of the spin system. The Bayesian formalism provides a global optimization strategy which allows to efficiently include prior knowledge about the spin system or to incorporate additional constraints concerning the parameters of interest. The proposed algorithm has been tested on synthetic and real 1D 1H-NMR data for various spin systems with increasing complexity. The results show that the Bayesian algorithm provides accurate estimates even for complex spectra with many overlapping regions, and that it can cope with symmetry induced local minima. By providing an unbiased estimate of the model evidence the proposed algorithm furthermore offers a way to discriminate between different spin system candidates.

Small molecules dominate organic phosphorus in NaOH-EDTA extracts of soils as determined by 31P NMR

Understanding the composition of organic phosphorus (P) in soils is relevant to various disciplines, from agricultural sciences to ecology. Despite past efforts, the precise nature of soil organic P remains an enigma, especially that of the orthophosphate monoesters, which dominate 31P NMR spectra of NaOH-EDTA extracts of soils worldwide. The monoester region often exhibits an unidentified, broad background believed to represent high molecular weight (MW) P. We investigated this monoester background using 1D 31P NMR and 2D 1H31P NMR, as well as 31P transverse relaxation (T2) measurements to calculate its intrinsic linewidth and relate it to MW. Analyzing seven soils from different ecosystems, we observed linewidths of 0.5 to 3 Hz for resolved monoester signals and the background, indicating that it consists of many, possibly >100, sharp signals associated with small (<1.5 kDa) organic P molecules. This result was further supported by 2D 1H31P NMR spectra revealing signals not resolved in the 1D spectra. Our findings align with 31P NMR studies detecting background signals in soil-free samples and modern evidence that alkali-soluble soil organic matter consists of self-assemblies of small organic compounds mimicking large molecules.

New analytical methods focusing on polar metabolite analysis in mass spectrometry and NMR-based metabolomics

Following in the footsteps of genomics and proteomics, metabolomics has revolutionised the way we investigate and understand biological systems. Rapid development in the last 25 years has been driven largely by technical innovations in mass spectrometry and nuclear magnetic resonance spectroscopy. However, despite the modest size of metabolomes relative to proteomes and genomes, methodological capabilities for robust, comprehensive metabolite analysis remain a major challenge. Therefore, development of new methods and techniques remains vital for progress in the field. Here, we review developments in LC-MS, GC-MS and NMR methods in the last few years that have enhanced quantitative and comprehensive metabolome coverage, highlighting the techniques involved, their technical capabilities, relative performance, and potential impact.

Deconvolution and Analysis of the 1H NMR Spectra of Crude Reaction Mixtures

Nuclear magnetic resonance (NMR) spectroscopy is an important analytical technique in synthetic organic chemistry, but its integration into high-throughput experimentation workflows has been limited by the necessity of manually analyzing the NMR spectra of new chemical entities. Current efforts to automate the analysis of NMR spectra rely on comparisons to databases of reported spectra for known compounds and, therefore, are incompatible with the exploration of new chemical space. By reframing the NMR spectrum of a reaction mixture as a joint probability distribution, we have used Hamiltonian Monte Carlo Markov Chain and density functional theory to fit the predicted NMR spectra to those of crude reaction mixtures. This approach enables the deconvolution and analysis of the spectra of mixtures of compounds without relying on reported spectra. The utility of our approach to analyze crude reaction mixtures is demonstrated with the experimental spectra of reactions that generate a mixture of isomers, such as Wittig olefination and C-H functionalization reactions. The correct identification of compounds in a reaction mixture and their relative concentrations is achieved with a mean absolute error as low as 1%.

Automated nuclear magnetic resonance fingerprinting of mixtures

Nuclear magnetic resonance (NMR) spectroscopy is a powerful tool for qualitative and quantitative analysis. However, for complex mixtures, determining the speciation from NMR spectra can be tedious and sometimes even unfeasible. On the other hand, identifying and quantifying structural groups in a mixture from NMR spectra is much easier than doing the same for components. We call this group-based approach "NMR fingerprinting." In this work, we show that NMR fingerprinting can even be performed in an automated way, without expert knowledge, based only on standard NMR spectra, namely, 13C, 1H, and 13C DEPT NMR spectra. Our approach is based on the machine-learning method of support vector classification (SVC), which was trained here on thousands of labeled pure-component NMR spectra from open-source data banks. We demonstrate the applicability of the automated NMR fingerprinting using test mixtures, of which spectra were taken using a simple benchtop NMR spectrometer. The results from the NMR fingerprinting agree remarkably well with the ground truth, which was known from the gravimetric preparation of the samples. To facilitate the application of the method, we provide an interactive website (https://nmr-fingerprinting.de), where spectral information can be uploaded and which returns the NMR fingerprint. The NMR fingerprinting can be used in many ways, for example, for process monitoring or thermodynamic modeling using group-contribution methods-or simply as a first step in species analysis.

Localized apelin-17 analogue-bicelle interactions as a facilitator of membrane-catalyzed receptor recognition and binding

The apelinergic system encompasses two peptide ligand families, apelin and apela, along with the apelin receptor (AR or APJ), a class A G-protein-coupled receptor. This system has diverse physiological effects, including modulating heart contraction, vasodilation/constriction, glucose regulation, and vascular development, with involvement in a variety of pathological conditions. Apelin peptides have been previously shown to interact with and become structured upon binding to anionic micelles, consistent with a membrane-catalyzed mechanism of ligand-receptor binding. To overcome the challenges of observing nuclear magnetic resonance (NMR) spectroscopy signals of a dilute peptide in biological environments, 19F NMR spectroscopy, including diffusion ordered spectroscopy (DOSY) and saturation transfer difference (STD) experiments, was used herein to explore the membrane-interactive behaviour of apelin. NMR-optimized apelin-17 analogues with 4-trifluoromethyl-phenylalanine at various positions were designed and tested for bioactivity through ERK activation in stably-AR transfected HEK 293 T cells. Far-UV circular dichroism (CD) spectropolarimetry and 19F NMR spectroscopy were used to compare the membrane interactions of these analogues with unlabelled apelin-17 in both zwitterionic/neutral and net-negative bicelle conditions. Each analogue binds to bicelles with relatively weak affinity (i.e., in fast exchange on the NMR timescale), with preferential interactions observed at the cationic residue-rich N-terminal and mid-length regions of the peptide leaving the C-terminal end unencumbered for receptor recognition, enabling a membrane-anchored fly-casting mechanism of peptide search for the receptor. In all, this study provides further insight into the membrane-interactive behaviour of an important bioactive peptide, demonstrating interactions and biophysical behaviour that cannot be neglected in therapeutic design.

Molecular Melodies: Unraveling the Hidden Harmonies of NMR Spectroscopy

This work explores the evolution of auditory analysis in NMR spectroscopy, tracing its journey from a supplementary tool to visual methods such as oscilloscopes, to a technique sidelined due to technological advancements. Despite its renaissance in the late 1990s with artistic and scientific applications, widespread adoption was hindered by the necessity for hardware modifications and reliance on specialized software. Addressing these barriers, this paper introduces a new feature in Mnova NMR software that facilitates the easy auditory interpretation of NMR signals. We discuss new applications of this tool, emphasizing its utility in aiding the identification of specific functional groups by auditory analysis of the spectrum's multiplets, such as distinguishing between aromatic, olefinic, or aliphatic protons, thereby enriching the interpretative capabilities of NMR data.

Simple and User-Friendly Methodology for Crystal Water Determination by Quantitative Proton NMR Spectroscopy in Deuterium Oxide

In drug research and development, knowledge of the precise structure of an active ingredient is crucial. However, it is equally important to know the water content of the drug molecule, particularly the number of crystal waters present in its structure. Such knowledge ensures the avoidance of drug dosage and formulation errors since the number of water molecules affects the physicochemical and pharmaceutical properties of the molecule. Several methods have been used for crystal water measurements of organic compounds, of which thermogravimetry and crystallography may be the most common ones. To the best of our knowledge, solution-state NMR spectroscopy has not been used for crystal water determination in deuterium oxide. Quantitative NMR (qNMR) method will be presented in the paper with a comparison of single-crystal X-ray diffraction and thermogravimetric analysis results. The qNMR method for water content measurement is straightforward, reproducible, and accurate, including measurement of 1H NMR spectrum before and after the addition of the analyte compound, and the result can be calculated after integration of the reference compound, analyte, and HDO signals using the given equation. In practical terms, there is no need for weighing the samples under study, which makes it simple and is a clear advantage to the current determination methods. In addition, the crystal structures of two model bisphosphonates used herein are reported: that of monopotassium etidronate dihydrate and monosodium zoledronate trihydrate.

Artificial intelligence-driven shimming for parallel high field nuclear magnetic resonance

Rapid drug development requires a high throughput screening technology. NMR could benefit from parallel detection but is hampered by technical obstacles. Detection sites must be magnetically shimmed to ppb uniformity, which for parallel detection is precluded by commercial shimming technology. Here we show that, by centering a separate shim system over each detector and employing deep learning to cope with overlapping non-orthogonal shimming fields, parallel detectors can be rapidly calibrated. Our implementation also reports the smallest NMR stripline detectors to date, based on an origami technique, facilitating further upscaling in the number of detection sites within the magnet bore.

Structure and dynamics of the mitochondrial DNA-compaction factor Abf2 from S. cerevisiae

Mitochondria are essential organelles that produce most of the energy via the oxidative phosphorylation (OXPHOS) system in all eukaryotic cells. Several essential subunits of the OXPHOS system are encoded by the mitochondrial genome (mtDNA) despite its small size. Defects in mtDNA maintenance and expression can lead to severe OXPHOS deficiencies and are amongst the most common causes of mitochondrial disease. The mtDNA is packaged as nucleoprotein structures, referred to as nucleoids. The conserved mitochondrial proteins, ARS-binding factor 2 (Abf2) in the budding yeast Saccharomyces cerevisiae (S. cerevisiae) and mitochondrial transcription factor A (TFAM) in mammals, are nucleoid associated proteins (NAPs) acting as condensing factors needed for packaging and maintenance of the mtDNA. Interestingly, gene knockout of Abf2 leads, in yeast, to the loss of mtDNA and respiratory growth, providing evidence for its crucial role. On a structural level, the condensing factors usually contain two DNA binding domains called high-mobility group boxes (HMG boxes). The co-operating mechanical activities of these domains are crucial in establishing the nucleoid architecture by bending the DNA strands. Here we used advanced solution NMR spectroscopy methods to characterize the dynamical properties of Abf2 together with its interaction with DNA. We find that the two HMG-domains react notably different to external cues like temperature and salt, indicating distinct functional properties. Biophysical characterizations show the pronounced difference of these domains upon DNA-binding, suggesting a refined picture of the Abf2 functional cycle.

A review of machine learning applications for the proton MR spectroscopy workflow

This literature review presents a comprehensive overview of machine learning (ML) applications in proton MR spectroscopy (MRS). As the use of ML techniques in MRS continues to grow, this review aims to provide the MRS community with a structured overview of the state-of-the-art methods. Specifically, we examine and summarize studies published between 2017 and 2023 from major journals in the MR field. We categorize these studies based on a typical MRS workflow, including data acquisition, processing, analysis, and artificial data generation. Our review reveals that ML in MRS is still in its early stages, with a primary focus on processing and analysis techniques, and less attention given to data acquisition. We also found that many studies use similar model architectures, with little comparison to alternative architectures. Additionally, the generation of artificial data is a crucial topic, with no consistent method for its generation. Furthermore, many studies demonstrate that artificial data suffers from generalization issues when tested on in vivo data. We also conclude that risks related to ML models should be addressed, particularly for clinical applications. Therefore, output uncertainty measures and model biases are critical to investigate. Nonetheless, the rapid development of ML in MRS and the promising results from the reviewed studies justify further research in this field.

Automatic classification of signal regions in 1H Nuclear Magnetic Resonance spectra

The identification and characterization of signal regions in Nuclear Magnetic Resonance (NMR) spectra is a challenging but crucial phase in the analysis and determination of complex chemical compounds. Here, we present a novel supervised deep learning approach to perform automatic detection and classification of multiplets in 1H NMR spectra. Our deep neural network was trained on a large number of synthetic spectra, with complete control over the features represented in the samples. We show that our model can detect signal regions effectively and minimize classification errors between different types of resonance patterns. We demonstrate that the network generalizes remarkably well on real experimental 1H NMR spectra.