Improvement in Signal-to-Noise Ratio of Liquid-State NMR Spectroscopy via a Deep Neural Network DN-Unet

Abstract

Nuclear magnetic resonance (NMR) is one of the most powerful analytical tools and is extensively applied in many fields. However, compared to other spectroscopic techniques, NMR has lower sensitivity, impeding its wider applications. Using data postprocessing techniques to increase the NMR spectral signal-to-noise ratio (SNR) is a relatively simple and cost-effective method. In this work, a deep neural network, termed as DN-Unet, is devised to suppress noise in liquid-state NMR spectra to enhance SNR. It combines structures of encoder–decoder and convolutional neural network. Different from traditional deep learning training strategy, M-to-S strategy is developed to enhance DN-Unet capability that multiple noisy spectra (inputs) correspond to a same single noiseless spectrum (label) in the training stage. The trained 1D model can be used for denoising not only 1D but also high dimension spectra, further improving DN-Unet’s performance. 1D, 2D, and 3D NMR spectra were utilized to evaluate DN-Unet performance. The results suggest that DN-Unet provides larger than 200-fold increase in SNR with weak peaks hidden in noise perfectly recovered and spurious peaks suppressed well. Since DN-Unet developed here to increase SNR is based on data postprocessing, it is universal for a variety of samples and NMR platforms. The great SNR enhancement and extreme excellence in differentiating signal and noise would greatly promote various liquid-state NMR applications.

The exposome paradigm to predict environmental health in terms of systemic homeostasis and resource balance based on NMR data science

“…234 DN-Unet combines structures of encoders-decoders, and CNN can be used to suppress noise in liquid-state NMR spectra to enhance SNR. 235 Lastly, NMR-TS automatically identies a molecule from its NMR spectrum, and discovers candidate molecules whose NMR spectra match the target spectrum by using RNN and density functional theory-computed spectra. 236 Several libraries have been constructed to ease implementation of ML, including MXNet (https://mxnet.apache.org/), PyTorch (https://pytorch.org/), Tensorow (https://www.tensorow.org/), scikit-multilearn (http://scikit.ml/), Keras (https://keras.io/), classyre, 237 MetNormalizer, 238 Caret 239 and mlr.…”

“…GPU computing has been applied to Monte Carlo simulation, 241,242 prediction of NMR chemical shis, 243,244 calculating the diffusion tensor for exible molecules, deep learning for metabolomics, 245 de novo pulse sequence design in solid-state NMR, 246 reconstruction of non-uniformly sampled NMR spectra 230,231 and denoising. 235,247 GPU computing is readily available through a workstation-class machine or cloud computing services (e.g., Amazon Web Service, Google Cloud Platform, Microso Azure).…”

“… 234 DN-Unet combines structures of encoders–decoders, and CNN can be used to suppress noise in liquid-state NMR spectra to enhance SNR. 235 Lastly, NMR-TS automatically identifies a molecule from its NMR spectrum, and discovers candidate molecules whose NMR spectra match the target spectrum by using RNN and density functional theory-computed spectra. 236 …”

Evaluation of the Higher Order Structure of Biotherapeutics Embedded in Hydrogels for Bioprinting and Drug Release

“…Nuclear magnetic resonance (NMR) and mass spectrometry are well-established techniques to investigate the preservation of the HOS of biologics in solution. 29 − 39 Solution NMR has been used previously on small proteins and peptides embedded in hydrogels to investigate the folding state in a confined environment 40 and for the structural characterization through residual dipolar couplings, since hydrogels behave as anisotropic external alignment media. 41 − 43 However, when the size of the pores in the gels is too small or strong interactions between the gel matrix and the cargo protein take place, the rotational correlation time of the protein in solution increases and makes solution NMR ineffective in the analysis of the protein structure at the atomic level.…”

SpinSPJ: a novel NMR scripting system to implement artificial intelligence and advanced applications

“…To support above usage scenarios and rapidly developing NMR applications, emerging technologies have to be quickly integrated in NMR software. For instance, deep learning [ 3 ] has been successfully applied in non-uniform sampling [ 1 ], spectrum denoising [ 4 ], chemical shift prediction [ 5 – 8 ], etc. Multivariate statistical analysis [ 2 ] plays an important role in metabolomics [ 9 , 10 ] to reveal the relationships between metabolites and significant issues such as diseases and biological processes.…”