October 11, 2007, ACD/Labs NJ NMR Seminar 2007, New Brunswick, NJ

Validating Compound Registrations with Automated NMR Verification in Open-Access

Phil Keyes

Abstract

Since the introduction of NMR prediction software, medicinal chemists have imagined submitting their compounds to corporate compound registration systems that would ultimately display a simplified pass/fail result. We have implemented a system where HPLC and LC/MS automated validation is embedded within our industry standard sample submission and supervisor verification process. By integrating the Gradient-HSQC 2D NMR experiment together with commercially available prediction software, we have added NMR to our verification process. This automated multidisciplinary approach, running in an open-access environment, acts as a means to assist in the review of final compounds submitted to our corporate compound registration system, and streamlines this process by identifying the "potentially" problematic compounds. This approach provides unprecedented availability of fully searchable analytical data throughout the life of a project. Results are rendered to a simplified overview in a graphic format that allows for rapid visualization of results. Results continue to be added, day by day, as we train, tune, and optimize our growing system to improve accuracy of results. Auxiliary benefits reaped from this process include automated generation of spectral databases, facilitation of NMR prediction database training assignments, and general archiving of data. The net result of this effort is helping to provide an objective measurement of compound collection quality, elimination of expensive downstream analyses performed on undesired structures, and reduction in data interpretation time by leveraging our computational assets thus freeing synthetic chemists to spend more time making molecules, analyzing SAR results, and focusing on discovering breakthrough treatments for human disease.

Download the presentation in Adobe Acrobat format (2.21 Mb PDF file).