Machine learning assistive rapid, label-free molecular phenotyping of blood with two-dimensional NMR correlational spectroscopy

Abstract

10Translation of the findings in basic science and clinical research into routine practice is hampered 11 by large variations in human phenotype. Developments in genotyping and phenotyping, such as 12 proteomics and lipidomics, are beginning to address these limitations. In this work, we developed 13 a new methodology for rapid, label-free molecular phenotyping of biological fluids (e.g., blood) 14 by exploiting the recent advances in fast and highly efficient multidimensional inverse Laplace 15 decomposition technique. We demonstrated that using two-dimensional T1-T2 correlational 16 spectroscopy on a single drop of blood (<5 μL), highly time-and patient-specific ´molecular 17 fingerprint´ can be obtained in minutes. Machine learning techniques were introduced to 18 transform the NMR correlational map into user-friendly information for point-of-care disease 19 diagnostic. The clinical utilities of this technique were demonstrated through the direct analysis 20 of human whole blood in various physiological (e.g., oxygenated/deoxygenated states) and 21 pathological (e.g., blood oxidation, hemoglobinopathies) conditions. (145 words) 22 23 Keywords: rapid phenotyping of blood, 2D NMR correlational spectroscopy, machine learning, 24 disease diagnostic and monitoring. Introduction 1High-resolution nuclear magnetic resonance (NMR) spectroscopy is a powerful and attractive 2 technique in biochemistry (e.g., for structural protein analysis 1 , characterizing metabolomics 3 responses in biological samples 2-4 ) and inorganic chemistry 5 . However, high-resolution NMR 4 systems are large, expensive and incompatible with in-situ or portable applications. There is an 5 increasing demand for low field portable NMR system for use in food sciences 6 , oil-gas 6 exploration 7 , and point of care clinical testing [8][9][10] . In high field NMR, biochemical information is 7 typically detected and encoded in the frequency domain ("chemical shift"), in which the spectral 8 resolution scale with respect to the external magnetic field. This reduces its portability and limit 9 its downstream application in a large scale manner. 10However, biochemical and biophysical information (e.g., molecular rotational, diffusional 11 motion) can also be encoded in the relaxation times frame, namely the longitudinal (T1) and 12 transverse (T2) using a portable low field NMR system. In addition, molecular information in the 13 time-domain can be inversely decoded with the availability of fast and reliable Laplace inversion 14 algorithm 7,11 . This can provide parallel information that is not available in the traditional NMR 15 frequency domain based spectra. 16In recent years, significant advances in NMR system miniaturization 8,12-14 (e.g., electronic 17 console 12,13,15,16 , radio-frequency probe 9,10,17-19 , microfluidic-based chip 20,21 ) utilizing small foot-18 print permanent magnetic (<1 Tesla) for one-dimensional NMR relaxometry on water-proton 19 (e.g., spin-spin relaxation (T2-relaxation)) have been widely applied for point-of-care medi...

Research in Metabolomics via Nuclear Magnetic Resonance Spectroscopy: Data Mining, Biochemistry and Clinical Chemistry

“…The results showed that ML approaches outperformed the human being and took a much shorter time. Therefore, the authors demonstrated the clinical efficacy of this technique by analyzing human blood in different physiological and pathological conditions, such as oxidation states [130]. Concerning the analysis of different physiological conditions, Figure 26 reports the T 1 -T 2 correlational maps of blood cells at oxygenated (a), oxidized (b), and deoxygenated (c) states.…”

“…Peng et al [130], by using two-dimensional NMR correlational spectroscopy on the longitudinal (T 1 ) and transversal components (T 2 ) of the magnetization relaxation time during its equilibrium recovery, were able to perform a molecular phenotyping of blood with the employment of supervised learning models, including neural networks. In detail, by means of a fast two-dimensional Laplace inversion [117], they obtained T 1 -T 2 correlation spectra on a single drop of blood (<5 μL) in a few minutes (Figure 26) with a benchtopsized NMR spectrometer.…”

Machine learning assistive rapid, label-free molecular phenotyping of blood with two-dimensional NMR correlational spectroscopy

“…We demonstrated (to the best of our knowledge) the first unique two-dimensional 'molecular fingerprint' of a single drop of blood (<5 µL) obtained in minutes using two dimensional T 1 -T 2 correlational spectroscopy with an inexpensive, benchtop-sized NMR spectrometer 28,29 . By exploiting the recent development of fast and highly efficient multidimensional inverse Laplace decomposition algorithm 7,30 , unique two-dimensional signature of various hemoglobin (Hb) derivatives with respect to its magnetic resonance relaxation reservoirs in oxygenated (oxy-Hb), deoxygenated (deoxy-Hb) and oxidized (oxidized Hb) states were observed for the first time (to the best of our knowledge) and its phenotypic expression in various pathological states (e.g., blood oxidation, hemoglobinopathies) are reported in this work.…”

Molecular phenotyping of oxidative stress in diabetes mellitus with point-of-care NMR system

“…To enumerate the various redox states of the Hb (e.g., Fe 2+ , Fe 3+ , Fe 4+ , and globin-associated radical Fe 4+ ) and the plasma, twodimensional relaxations map, known as T 1 -T 2 magnetic state diagram was proposed ( Fig. 1d) 34,35 . This magnetic state diagram allows visualization and identification of the intermediate redox states and the transient, dynamic pathways of the blood sample.…”

“…Recent advances in NMR system miniaturization have raised the prospect of applying these techniques in point-ofcare diagnostic setting [36][37][38] . These applications include immunomagnetic labeling based detection (e.g., tumor cells [39][40][41] , tuberculosis 42 , and magneto-DNA detection of bacteria 43 ) and label-free micro MR detection of various diseases (e.g., oxygenation 44 /oxidation 31 level of the blood, malaria screening 30,33,45 , diabetes mellitus 28 , and hemoglobinopathies 35 ).…”