scholarly journals A General Protocol for the Accurate Predictions of Molecular 13C/1H NMR Chemical Shifts via Machine Learning

2019 ◽  
Author(s):  
Peng Gao ◽  
Jun Zhang ◽  
Qian Peng ◽  
Vassiliki-Alexandra Glezakou
Keyword(s):  
Machine Learning ◽  
1H Nmr ◽  
Density Functional ◽  
Organic Molecules ◽  
Chemical Shifts ◽  
Atomic Orbital ◽  
Computational Cost ◽  
Experimental Studies ◽  
Dft Methods ◽  
Nmr Chemical Shifts

Accurate prediction of NMR chemical shifts with affordable computational cost is of great importance for rigorous structural assignments of experimental studies. However, the most popular computational schemes for NMR calculation—based on density functional theory (DFT) and gauge-including atomic orbital (GIAO) methods—still suffer from ambiguities in structural assignments. Using state-of-the-art machine learning (ML) techniques, we have developed a DFT+ML model that is capable of predicting 13C/1H NMR chemical shifts of organic molecules with high accuracy. The input for this generalizable DFT+ML model contains two critical parts: one is a vector providing insights into chemical environments, which can be evaluated without knowing the exact geometry of the molecule; the other one is the DFT-calculated isotropic shielding constant. The DFT+ML model was trained with a dataset containing 476 13C and 270 1H experimental chemical shifts. For the DFT methods used here, the root-mean-square-derivations (RMSDs) for the errors between predicted and experimental 13C/1H chemical shifts are as small as 2.10/0.18 ppm, which is much lower than the typical DFT (5.54/0.25 ppm), or DFT+linear regression (4.77/0.23 ppm) approaches. It also has smaller RMSDs and maximum absolute errors than two previously reported NMR-predicting ML models. We test the robustness of the model on two classes of organic molecules (TIC10 and hyacinthacines), where we unambiguously assigned the correct isomers to the experimental ones. This DFT+ML model is a promising way of predicting NMR chemical shifts and can be easily adapted to calculated shifts for any chemical compound.<br>

scholarly journals A General Protocol for the Accurate Predictions of Molecular 13C/1H NMR Chemical Shifts via Machine Learning

2019 ◽  
Author(s):  
Peng Gao ◽  
Jun Zhang ◽  
Qian Peng ◽  
Vassiliki-Alexandra Glezakou
Keyword(s):  
Machine Learning ◽  
1H Nmr ◽  
Density Functional ◽  
Organic Molecules ◽  
Chemical Shifts ◽  
Atomic Orbital ◽  
Computational Cost ◽  
Experimental Studies ◽  
Dft Methods ◽  
Nmr Chemical Shifts

Accurate prediction of NMR chemical shifts with affordable computational cost is of great importance for rigorous structural assignments of experimental studies. However, the most popular computational schemes for NMR calculation—based on density functional theory (DFT) and gauge-including atomic orbital (GIAO) methods—still suffer from ambiguities in structural assignments. Using state-of-the-art machine learning (ML) techniques, we have developed a DFT+ML model that is capable of predicting 13C/1H NMR chemical shifts of organic molecules with high accuracy. The input for this generalizable DFT+ML model contains two critical parts: one is a vector providing insights into chemical environments, which can be evaluated without knowing the exact geometry of the molecule; the other one is the DFT-calculated isotropic shielding constant. The DFT+ML model was trained with a dataset containing 476 13C and 270 1H experimental chemical shifts. For the DFT methods used here, the root-mean-square-derivations (RMSDs) for the errors between predicted and experimental 13C/1H chemical shifts are as small as 2.10/0.18 ppm, which is much lower than the typical DFT (5.54/0.25 ppm), or DFT+linear regression (4.77/0.23 ppm) approaches. It also has smaller RMSDs and maximum absolute errors than two previously reported NMR-predicting ML models. We test the robustness of the model on two classes of organic molecules (TIC10 and hyacinthacines), where we unambiguously assigned the correct isomers to the experimental ones. This DFT+ML model is a promising way of predicting NMR chemical shifts and can be easily adapted to calculated shifts for any chemical compound.<br>

BAND NN: A Deep Learning Framework For Energy Prediction and Geometry Optimization of Organic Small Molecules

2019 ◽  
Author(s):  
Siddhartha Laghuvarapu ◽  
Yashaswi Pathak ◽  
U. Deva Priyakumar
Keyword(s):  
Machine Learning ◽  
Density Functional ◽  
Computational Cost ◽  
Geometry Optimization ◽  
Dft Methods ◽  
Energy Prediction ◽  
Machine Learning Model ◽  
Equilibrium Structures ◽  
High Level ◽  
Non Equilibrium

Recent advances in artificial intelligence along with development of large datasets of energies calculated using quantum mechanical (QM)/density functional theory (DFT) methods have enabled prediction of accurate molecular energies at reasonably low computational cost. However, machine learning models that have been reported so far requires the atomic positions obtained from geometry optimizations using high level QM/DFT methods as input in order to predict the energies, and do not allow for geometry optimization. In this paper, a transferable and molecule-size independent machine learning model (BAND NN) based on a chemically intuitive representation inspired by molecular mechanics force fields is presented. The model predicts the atomization energies of equilibrium and non-equilibrium structures as sum of energy contributions from bonds (B), angles (A), nonbonds (N) and dihedrals (D) at remarkable accuracy. The robustness of the proposed model is further validated by calculations that span over the conformational, configurational and reaction space. The transferability of this model on systems larger than the ones in the dataset is demonstrated by performing calculations on select large molecules. Importantly, employing the BAND NN model, it is possible to perform geometry optimizations starting from non-equilibrium structures along with predicting their energies.

scholarly journals General QSPR Protocol for Atomic/Inter-atomic Properties Predictions: Fragments based Graph Convolutional Neural Network (F-GCN)

2021 ◽  
Author(s):  
Peng Gao ◽  
Jie Zhang ◽  
Hongbo Qiu ◽  
Shuaifei Zhao
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Density Functional ◽  
Chemical Shifts ◽  
Computational Cost ◽  
Accurate Solution ◽  
Structure Property ◽  
Nmr Chemical Shifts ◽  
H Nmr ◽  
Atomic Properties

In this study, a general quantitative structure-property relationship (QSPR) protocol, fragments based graph convolutional neural network (F-GCN), was developed for atomic and inter-atomic properties predictions. We applied this novel artificial intelligence (AI) tool in NMR chemical shifts and bond dissociation energies (BDEs) predictions. The predicted results were comparable to experimental measurement, while the computational cost was substantially reduced, with respect to pure density functional theory (DFT) calculations. The two important features of F-GCN can be summarised as: first, it could utilise different levels of molecular fragments centered at the target chemical bonds for atomic and inter-atomic information extraction; second, the designed architecture is also open to include additional descriptors for more accurate solution of chemical environment, making itself more efficient for local properties descriptions. And during our test, the averaged prediction error of <sup>1</sup>H NMR chemical shifts can be as small as 0.32 ppm; and the error of C-H BDEs estimations, is 2.7 kcal/mol. Moreover, we further demonstrated the applicability of this developed F-GCN model via several challenging structural assignments. The success of the F-GCN in atomic and inter-atomic predictions also indicates an essential improvement of computational chemistry with the assistance of AI tools.

scholarly journals Structure Analysis of Sulfated Polysaccharides Extracted from Scinaia interrupta: A Experimental and Density Functional Theory Studies

2020 ◽  
Vol 32 (7) ◽  
pp. 1589-1596
Author(s):  
Nivedita Acharjee ◽  
Tuhin Ghosh
Keyword(s):  
1H Nmr ◽  
Red Algae ◽  
Density Functional ◽  
Chemical Shifts ◽  
Present Report ◽  
Monomer Unit ◽  
Absolute Error ◽  
Sulfated Polysaccharides ◽  
Nmr Studies ◽  
Nmr Chemical Shifts

In present report, a combined experimental and theoretical study has been performed to address the isolation procedure and spectroscopic structure elucidation of polysaccharides such as xylomannan isolated from marine red algal source Scinaia interrupta. The structure of the polysaccharides obtained from the red algae of Scinaia interrupta has been studied from NMR, IR and GC-MS spectroscopy. The investigation revealed that red algae contained a backbone of α-(1→4)-linked D-mannopyranosyl residues substituted at 6-position with a single stub of β-D-xylopyranosyl residues. The major polysaccharide, which had 0.6 sulfate groups per monomer unit and an apparent molecular mass of 120 KDa. The backbone structure was optimized at DFT/B3LYP/6-311G(d,p) level of theory and GIAO-NMR studies were performed at B3LYP/6-311++G(2d,p) level of theory followed by mean absolute error calculations of the computed chemical shifts for two possible conformers resulting from the flipping of xylopyranosyl residue. The NMR calculations were in agreement with the experimental findings. The experimental 1H NMR chemical shifts were then correlated with the NBO, Merz Kollman (MK), ChelpG and Mulliken charges of the predicted conformer. A reasonable correlation with the experimental 1H NMR chemical shifts and the computed NBO charges with correlation coefficient of 0.906.

Optimized Structures and Theoretical 1H-NMR Spectral Analysis of the Methylene Protons in Dibenzyl Sulfoxide

2011 ◽  
Vol 391-392 ◽  
pp. 1368-1374 ◽  
Author(s):  
Zheng Ping Wu ◽  
Yuan Bing Sun ◽  
Ian S. Butler
Keyword(s):  
1H Nmr ◽  
Density Functional ◽  
Chemical Shifts ◽  
Dft Methods ◽  
Functional Theory ◽  
H Nmr ◽  
Effect Of Solvent ◽  
Solvent Shift ◽  
Total Energies ◽  
Good Agreement

Dibenzyl sulfoxide [C6H5CH2)2SO, DBzSO] has been studied using density functional theory (DFT) methods with a particular emphasis on the theoretical 1H-NMR spectra of the methylene protons. The 1H-NMR chemical shifts of the methylene protons of DBzSO can be divided into two main types. Four possible structures of DBzSO were considered and the total energies were calculated for both a vacuum and in CDCl3 solvent. The change of length of S-O and S-C bonds in solvent was more obvious than that of the C(CH2)-C(C6H5) bonds; The S-O bond was longer and S-C bond was shorter in CDCl3. The essence effect of solvent on the properties of dibenzyl sulfoxide should come from the change of the geometrical structure. The change of shift Δx, [shift (solvent) - shift (vacuum)] showed that the effect of solvent on methylene protons of dibenzyl sulfoxide was apparent. Except of the other H of the rings, the two ortho H which were near S-O bond appeared more sensitivity on the solvent. The optimized structures in CDCl3 were in good agreement with the experimental data. The NMR peaks of methylene protons should be split more apparently in actual circumstance and the complex split of CH2 1HNMR peaks should be explained in some degree.

scholarly journals General Protocol for the Accurate Prediction of Molecular 13C/1H NMR Chemical Shifts via Machine Learning Augmented DFT

2020 ◽  
Vol 60 (8) ◽  
pp. 3746-3754 ◽  
Author(s):  
Peng Gao ◽  
Jun Zhang ◽  
Qian Peng ◽  
Jie Zhang ◽  
Vassiliki-Alexandra Glezakou
Keyword(s):  
Machine Learning ◽  
1H Nmr ◽  
Chemical Shifts ◽  
Accurate Prediction ◽  
Nmr Chemical Shifts

BAND NN: A Deep Learning Framework For Energy Prediction and Geometry Optimization of Organic Small Molecules

2019 ◽  
Author(s):  
Siddhartha Laghuvarapu ◽  
Yashaswi Pathak ◽  
U. Deva Priyakumar
Keyword(s):  
Machine Learning ◽  
Density Functional ◽  
Computational Cost ◽  
Geometry Optimization ◽  
Dft Methods ◽  
Energy Prediction ◽  
Machine Learning Model ◽  
Equilibrium Structures ◽  
High Level ◽  
Non Equilibrium

Recent advances in artificial intelligence along with development of large datasets of energies calculated using quantum mechanical (QM)/density functional theory (DFT) methods have enabled prediction of accurate molecular energies at reasonably low computational cost. However, machine learning models that have been reported so far requires the atomic positions obtained from geometry optimizations using high level QM/DFT methods as input in order to predict the energies, and do not allow for geometry optimization. In this paper, a transferable and molecule-size independent machine learning model (BAND NN) based on a chemically intuitive representation inspired by molecular mechanics force fields is presented. The model predicts the atomization energies of equilibrium and non-equilibrium structures as sum of energy contributions from bonds (B), angles (A), nonbonds (N) and dihedrals (D) at remarkable accuracy. The robustness of the proposed model is further validated by calculations that span over the conformational, configurational and reaction space. The transferability of this model on systems larger than the ones in the dataset is demonstrated by performing calculations on select large molecules. Importantly, employing the BAND NN model, it is possible to perform geometry optimizations starting from non-equilibrium structures along with predicting their energies.

scholarly journals General QSPR Protocol for Atomic/Inter-atomic Properties Predictions: Fragments based Graph Convolutional Neural Network (F-GCN)

2021 ◽  
Author(s):  
Peng Gao ◽  
Jie Zhang ◽  
Hongbo Qiu ◽  
Shuaifei Zhao
Keyword(s):  
Neural Network ◽  
Convolutional Neural Network ◽  
Density Functional ◽  
Chemical Shifts ◽  
Computational Cost ◽  
Accurate Solution ◽  
Structure Property ◽  
Nmr Chemical Shifts ◽  
H Nmr ◽  
Atomic Properties

In this study, a general quantitative structure-property relationship (QSPR) protocol, fragments based graph convolutional neural network (F-GCN), was developed for atomic and inter-atomic properties predictions. We applied this novel artificial intelligence (AI) tool in NMR chemical shifts and bond dissociation energies (BDEs) predictions. The predicted results were comparable to experimental measurement, while the computational cost was substantially reduced, with respect to pure density functional theory (DFT) calculations. The two important features of F-GCN can be summarised as: first, it could utilise different levels of molecular fragments centered at the target chemical bonds for atomic and inter-atomic information extraction; second, the designed architecture is also open to include additional descriptors for more accurate solution of chemical environment, making itself more efficient for local properties descriptions. And during our test, the averaged prediction error of <sup>1</sup>H NMR chemical shifts can be as small as 0.32 ppm; and the error of C-H BDEs estimations, is 2.7 kcal/mol. Moreover, we further demonstrated the applicability of this developed F-GCN model via several challenging structural assignments. The success of the F-GCN in atomic and inter-atomic predictions also indicates an essential improvement of computational chemistry with the assistance of AI tools.

scholarly journals Quantum chemical studies on structural, spectroscopic, nonlinear optical, and thermodynamic properties of the 1,2,4-triazole compound

2021 ◽  
Vol 27 (1) ◽  
pp. 112-132
Author(s):  
Hilal Medetalibeyoğlu ◽  
Haydar Yüksek
Keyword(s):  
Density Functional Theory ◽  
Molecular Orbital ◽  
Density Functional ◽  
Chemical Shifts ◽  
Atomic Orbital ◽  
Fluorescence Emission ◽  
Strong Relationship ◽  
Basis Set ◽  
Functional Theory ◽  
Nmr Chemical Shifts

Abstract In this study, the structure of 4-[4-(diethylamino)-benzylideneamino]-5-benzyl-2H-1,2,4-triazol-3(4H)-one (DBT) was examined through spectroscopic and theoretical analyses. In this respect, the geometrical, vibrational frequency, 1H and 13C-nuclear magnetic resonance (NMR) chemical shifts, thermodynamic, hyperpolarizability, and electronic properties including the highest occupied molecular orbital–lowest unoccupied molecular orbital (HOMO–LUMO) energies of DBT as a potential non-linear optical (NLO) material were investigated using density functional theory at the B3LYP level with the 6-311G basis set. 1H and 13C-NMR chemical shifts of DBT with the gauge-invariant atomic orbital and continuous set of gauge transformation methods (in the solvents) were estimated, and the computed chemical shift values displayed excellent alignment with observed ones. Time-dependent density-functional theory (TD-DFT) calculations with the integral equation formalism polarizable continuum model within various solvents and gas phases in the ground state were used to evaluate UV-vis absorption and fluorescence emission wavelengths. Thermodynamic parameters including enthalpy, heat capacity, and entropy for DBT were also calculated at various temperatures. Moreover, calculations of the NLO were carried out to obtain the title compound’s electric dipole moment and polarizability properties. To illustrate the effect of the theoretical method on the spectroscopic and structural properties of DBT, experimental data of structural and spectroscopic parameters were used. The correlational analysis results were observed to indicate a strong relationship between the experimental and theoretical results.

Sign in / Sign up

Export Citation Format

Share Document