Statistical removal of background signals from high-throughput ¹H NMR line-broadening ligand-affinity screens

NMR ligand-affinity screens are vital to drug discovery, are routinely used to screen fragment-based libraries, and used to verify chemical leads from high-throughput assays and virtual screens. NMR ligand-affinity screens are also a highly informative first step towards identifying functional epitopes of unknown proteins, as well as elucidating the biochemical functions of protein-ligand interaction at their binding interfaces. While simple one-dimensional ¹H NMR experiments are capable of indicating binding through a change in ligand line shape, they are plagued by broad, ill-defined background signals from protein ¹H resonances. We present an uncomplicated method for subtraction of protein background in high-throughput ligand-based affinity screens, and show that its performance is maximized when phase-scatter correction is applied prior to subtraction.