Secondary structure of proteins from NMR data by neural nets.

Abstract The information on the secondary structure of a protein may be used both in its own right to determine properties of a protein and as an intermediate step towards the assessment of its tertiary structure. NMR spectroscopy has been used for proteins for 10-15 years only, but has already provided descriptions of proteins. This paper presents a new method for determining the seconay structure of a protein from its NMR (nuclear magnetic resonance) spectra. We present a neural network approach to determine the secondary structure of an amido acid.