TRD3 NMRbox: Bayesian Analytics

University Of Connecticut Sch Of Med/Dnt, Farmington CT

Investigators

Linked publications & trials

Abstract

Abstract: TRD 3. Biological applications of NMR have expanded to include broad areas of investigation in structural biology, metabolic studies, disease diagnosis, and drug discovery. NMR offers detailed information, but the interpretation involves a multistep and intricate process that uses a range of mathematical and computational methods. Powerful software packages have been developed to facilitate data collection, analysis, and interpretation, but significant challenges remain. The aim of this TRD is to deliver a robust and reliable statistical framework, based on Bayesian principles, in order to advance the state of NMR data analytics, simplify the future development of advanced NMR applications that use statistical principles, and provide significant new applications of this framework in the context of NMR. Central components of this TRD include the development of a Bayesian inference engine and a set of programming interfaces, the redeployment of some legacy applications on the new platform, provide templates for development of applications by other research teams, and integration of the technology within the platform being developed by the Center. The implementation of this TRD will benefit from key advantages offered by the NMRbox platform. The resulting Bayesian inference technology will see significantly lowered barriers to use, substantially reduced time for development of new applications, broader accessibility, and higher visibility. By facilitating the deployment, utilization, interoperation, and persistence of this advanced technology, the proposed TRD will advance the application of NMR for a wide range of challenging applications in biomedicine, and help ensure the reproducibility bio-NMR studies.