Estimating 1H NMR coupling constants with ANN models for chemical shifts: Spectra simulation in the SPINUS system
Joao Aires-de-Sousa and Yuri Binev. REQUIMTE and Department of Chemistry, New University of Lisbon, campus FCTUNL, 2829-516 Caparica, Portugal
Fast and accurate predictions of 1H NMR spectra of organic compounds play an important role in automatic structure elucidation and validation. The SPINUS program is a feed-forward neural network (FFNN) system developed over the last eight years for the prediction of 1H NMR properties from the molecular structure. It was trained using a series of empirical proton descriptors. The FFNNs were incorporated into Associative Neural Networks (ASNN), which correct a prediction obtained by the FFNNs with the observed errors for the k nearest neighbours in an additional memory. Here we show a procedure to estimate coupling constants with the ASNNs trained for chemical shifts. Now a memory of coupled protons and the experimental coupling constants is used. The ASNNs find the pairs of coupled protons most similar to a query, and these are used to estimate coupling constants. A web interface for 1H NMR spectra prediction is presented.