An official website of the United States government
Background: Quantification of metabolites from nuclear magnetic resonance (NMR) spectra in an accurate, high-throughput manner requires effective data processing tools. Neural networks are relatively underexplored in quantitative NMR metabolomics despite impressive speed and throughput compared to more conventional peak-fitting metabolomics software. Methods: This work investigates practices for dataset and model development in the task of metabolite quantification directly from simulated NMR spectra for three neural network models: the multi-layered perceptron, the convolutional neural network, and the transformer. Model architectures, training parameters, and training datasets are optimized before comparing each model on simulated 400-MHz 1H-NMR spectra of complex mixtures with 8, 44, or 86 metabolites to quantify in spectra ranging from simple to highly complex and overlapping peaks. The optimized models were further validated on spectra at 100- and 800-MHz. Results: The transformer was the most effective network for NMR metabolite quantification, especially as the number of metabolites per spectra increased or target concentrations were low or had a large dynamic range. Further, the transformer was able to accurately quantify metabolites in simulated spectra from 100-MHz up to 800-MHz. Conclusions: The methods developed in this work reveal that transformers have the potential to accurately perform fully automated metabolite quantification in real-time and, with further development with experimental data, could be the basis for automated quantitative NMR metabolomics software.
more »
« less
Neural Networks for Conversion of Simulated NMR Spectra from Low-Field to High-Field for Quantitative Metabolomics
Background: The introduction of benchtop NMR instruments has made NMR spectroscopy a more accessible, affordable option for research and industry, but the lower spectral resolution and SNR of a signal acquired on low magnetic field spectrometers may complicate the quantitative analysis of spectra. Methods: In this work, we compare the performance of multiple neural network architectures in the task of converting simulated 100 MHz NMR spectra to 400 MHz with the goal of improving the quality of the low-field spectra for analyte quantification. Multi-layered perceptron networks are also used to directly quantify metabolites in simulated 100 and 400 MHz spectra for comparison. Results: The transformer network was the only architecture in this study capable of reliably converting the low-field NMR spectra to high-field spectra in mixtures of 21 and 87 metabolites. Multi-layered perceptron-based metabolite quantification was slightly more accurate when directly processing the low-field spectra compared to high-field converted spectra, which, at least for the current study, precludes the need for low-to-high-field spectral conversion; however, this comparison of low and high-field quantification necessitates further research, comparison, and experimental validation. Conclusions: The transformer method of NMR data processing was effective in converting low-field simulated spectra to high-field for metabolomic applications and could be further explored to automate processing in other areas of NMR spectroscopy.
more »
« less
- PAR ID:
- 10586480
- Publisher / Repository:
- MDPI
- Date Published:
- Journal Name:
- Metabolites
- Volume:
- 14
- Issue:
- 12
- ISSN:
- 2218-1989
- Page Range / eLocation ID:
- 666
- Subject(s) / Keyword(s):
- NMR spectroscopy low-field NMR metabolomics neural networks transformer
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Two-dimensional nuclear magnetic resonance (2D NMR) spectroscopy was evaluated for the identification and quantification of compounds in an unknown street drug sample. Using 2D COSY and HSQC techniques, heroin was successfully quantified, and the presence of 6-monoacetylmorphine (6-MAM), xylazine, and caffeine was confirmed through partial structural elucidation. These methods demonstrated the ability to differentiate structurally similar opioid analogues without reliance on reference library databases. While gas chromatography–mass spectrometry (GC–MS) remains the standard in forensic laboratories, it has limitations in de novo structural analysis and in detecting emerging analogues absent from spectral libraries. In this study, heroin and fentanyl were quantified in both simulated and actual street samples at concentrations ranging from 0.97 to 1.80 mg/mL, with errors between 0% and 34% using a 400 MHz NMR instrument. A benchtop 60 MHz NMR system also detected and quantified 56 mg/mL of heroin with a 24% error in a simulated sample. These findings support the complementary role of 2D NMR spectroscopy in forensic drug analysis in light of the opioid epidemic and the evolving drug market.more » « less
-
Rapid and automated lipid profiling by nuclear magnetic resonance spectroscopy using neural networksAbstract Nuclear magnetic resonance (NMR) spectroscopy is a powerful tool for quantitative metabolomics; however, quantification of metabolites from NMR data is often a slow and tedious process requiring user input and expertise. In this study, we propose a neural network approach for rapid, automated lipid identification and quantification from NMR data. Multilayered perceptron (MLP) networks were developed with NMR spectra as the input and lipid concentrations as output. Three large synthetic datasets were generated, each with 55,000 spectra from an original 30 scans of reference standards, by using linear combinations of standards and simulating experimental‐like modifications (line broadening, noise, peak shifts, baseline shifts) and common interference signals (water, tetramethylsilane, extraction solvent), and were used to train MLPs for robust prediction of lipid concentrations. The performances of MLPS were first validated on various synthetic datasets to assess the effect of incorporating different modifications on their accuracy. The MLPs were then evaluated on experimentally acquired data from complex lipid mixtures. The MLP‐derived lipid concentrations showed high correlations and slopes close to unity for most of the quantified lipid metabolites in experimental mixtures compared with ground‐truth concentrations. The most accurate, robust MLP was used to profile lipids in lipophilic hepatic extracts from a rat metabolomics study. The MLP lipid results analyzed by two‐way ANOVA for dietary and sex differences were similar to those obtained with a conventional NMR quantification method. In conclusion, this study demonstrates the potential and feasibility of a neural network approach for improving speed and automation in NMR lipid profiling and this approach can be easily tailored to other quantitative, targeted spectroscopic analyses in academia or industry.more » « less
-
null (Ed.)Nuclear magnetic resonance (NMR) spectroscopy is a well-established analytical technique used to study chemicals and their transformations. However, high-eld NMR spectroscopy necessitates advanced infrastructure and even cryogen-free benchtop NMR spectrometers cannot be readily assembled from commercially available components. We demonstrate the construction of a portable zero-field NMR spectrometer employing a commercially available magnetometer and investigate its applications in analytical chemistry. In particular, J-spectra of small representative biomolecules [13C]-formic acid, [1-13C]-glycine, [2,3-13C]-fumarate, and [1-13C]-D-glucose were acquired and an approach relying on the presence of a transverse magnetic eld during the detection was investigated for relaxometry purposes. We found that water relaxation time strongly depends on the concentration of dissolved D-glucose in the range of 1-10 mM suggesting opportunities for indirect assessment of glucose concentration in aqueous solutions. Extending analytical capabilities of zero-field NMR to aqueous solutions of simple biomolecules (aminoacids, sugars and metabolites) and relaxation studies of aqueous solutions of glucose highlight the analytical potential of non-invasive and portable ZULF NMR sensors for applications outside of research laboratories.more » « less
-
Abstract Field‐stepped NMR spectroscopy at up to 36 T using the series‐connected hybrid (SCH) magnet at the U.S. National High Magnetic Field Laboratory is demonstrated for acquiring ultra‐wideline powder spectra of nuclei with very large quadrupolar interactions. Historically, NMR evolved from the continuous‐wave (cw) field‐swept method in the early days to the pulsed Fourier‐transform method in the modern era. Spectra acquired using field sweeping are generally considered to be equivalent to those acquired using the pulsed method. Here, it is shown that field‐stepped wideline spectra of half‐integer spin quadrupolar nuclei acquired using WURST/CPMG methods can be significantly different from those acquired with the frequency‐stepped method commonly used with superconducting magnets. The inequivalence arises from magnetic field‐dependent NMR interactions such as the anisotropic chemical shift and second‐order quadrupolar interactions; the latter is often the main interaction leading to ultra‐wideline powder patterns of half‐integer spin quadrupolar nuclei. This inequivalence needs be taken into account to accurately and correctly determine the quadrupolar coupling and chemical shift parameters. A simulation protocol is developed for spectral fitting to facilitate analysis of field‐stepped ultra‐wideline NMR spectra acquired using powered magnets. A MATLAB program which implements this protocol is available on request.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3390/metabo14120666
