An Artificial Neural Network (ANN) to Identify Organic and Bio Molecules Based On Their Nuclear Magnetic Resonance (NMR) Spectra

This paper reports a novel design and implementation of an Artificial Neural Network (ANN) aimed at recognizing unknown organic or biochemical compounds for which their Nuclear Magnetic Resonance (NMR) Spectra have been previously obtained. The purpose of developing this computational tool is to save time in chemistry and proteomics research. A chemist who needs to identify an unknown molecule for which its NMR spectrum has been recorded, can make use of this automated tool to retrieve the name and chemical structure of the closest known molecule matching the given NMR spectrum. This way, a researcher does not have to browse through large catalogs of NMR spectra to identify an unknown sample.