Non-linear modelling of 13C NMR chemical shift data using artificial neural networks and partial least squares method

13C NMR Chemical Shift Prediction of the sp3 Carbon Atoms in the α Position Relative to the Double Bond in Acyclic Alkenes

“…The growing interest in the application of Artificial Neural Networks (ANNs) in the field of computer-assisted spectral interpretation is a result of their demonstrated superiority over the traditional models . The use of ANNs in spectra interpretation and structure elucidation is 2-fold, i.e., either for classification (recognition of structural characteristics from IR − or MS spectra, − joint IR- 13 C-NMR spectra 1 or IR-MS spectra) or for a quantitative prediction of a certain atomic property (the chemical shift in 13 C NMR spectra). − …”

“…1 The use of ANNs in spectra interpretation and structure elucidation is 2-fold, i.e., either for classification (recognition of structural characteristics from IR [2][3][4][5][6][7][8][9][10][11] or MS spectra, [12][13][14][15] joint IR- 13 C-NMR spectra 1 or IR-MS spectra 16 ) or for a quantitative prediction of a certain atomic property (the chemical shift in 13 C NMR spectra). [17][18][19][20][21][22][23][24][25][26][27][28] In a previous paper 28 we have estimated the 13 C NMR chemical shift of sp 2 carbon atoms in acyclic alkenes with MultiLinear Regression (MLR) and MultiLayer Feedforward (MLF) ANN models, using as structural descriptor of the environment of the resonating carbon a Topo-Stereochemical Code (TSC) with 12 components allowing for a unique description of the topo-stereochemical location of the carbon atoms around the double bond. The study investigated the 13 C NMR chemical shift of 130 acyclic alkenes with 244 structurally unique sp 2 carbon atoms.…”

13C NMR Chemical Shift Prediction of sp2 Carbon Atoms in Acyclic Alkenes Using Neural Networks

“…The methods used in predicting 13 C NMR shifts are diverse, ranging from database retrieval , to additive relationships − and empirical models which include various topological, molecular, and quantum mechanics descriptors . Recent research efforts are directed toward the enhancement of the NMR shift predictions by nonlinear models, such as the neural network model applied to 13 C in alkanes and cycloalkanes, − in monosubstituted benzenes, , for keto-steroids, and halomethanes as well as for 31 P. , …”

“…The methods used in predicting 13 C NMR shifts are diverse, ranging from database retrieval 1,2 to additive relationships [3][4][5][6][7][8][9] and empirical models which include various topological, 10 molecular, and quantum mechanics descriptors. 11 Recent research efforts are directed toward the enhancement of the NMR shift predictions by nonlinear models, such as the neural network model applied to 13 C in alkanes and cycloalkanes, [12][13][14][15] in monosubstituted benzenes, 16,17 for keto-steroids, 18 and halomethanes 19 as well as for 31 P. 20,21 MultiLayer Feedforward (MLF) Artificial Neural Networks (ANN) 22,23 are a promising model for solving Quantitative Structure-Property Relationships (QSPR) problems, and they are particularly useful in cases where it is difficult to specify an exact mathematical model which describes a specific structure-property relationship. In such cases ANNs, which employ learning procedures based on the patterns describing the molecular structure and the investigated property in order to develop an internal representation of a physicochemical phenomena, may be able to form structural correlations which produce accurate predictions.…”