Improved Processing of Selective NMR Spectra of Biopolymers by Separation of Noise and Signal Subspaces through Singular-Value Decomposition

Abstract

A novel sensitivity-enhancement procedure is introduced for 2D NMR data matrices. It is based on the separation of the signal-and-noise subspaces by means of singular-value decomposition of the experimental 2D array. Although no theoretical limitation exists for a general application of the method, the reliability of the results increases considerably with reduced data sets such as those of selective measurements. Advantageous applications can be envisaged for the quantitation of NMR parameters in biopolymers whose linewidths are often large enough to undermine severely the sensitivity of selective experiments.