Computational Analyses of NMR Spectra

Building on the DP4 method, developed in the Goodman Group, the project will further extend the computational analysis of NMR spectra, so that more information may be gathered more easily from analytical data. This will be achieved in a number of different ways:
(1) Automated transfer of data from spectrometers to the analysis program. In the current version of the program, NMR spectra need to be analysed by hand, and the data prepared in a way suitable for publication in standard journals. This process is familiar, but time consuming. Algorithms are being developed so that this step can be eliminated, and the output of NMR machines can be piped directly to the DP4 algorithm.
(2) Currently, DP4 is limited to 1D spectra, even though 2D spectra are routinely acquired. Expanding the range of the method to 2D spectra will make it possible to extract even more knowledge from the NMR experiments that have already been run, and also guide experimentalists towards the most cost-effective choices for further NMR measurements.
(3) The bottlenecks in the method include the time required for conformation searching and finding energy minima for low energy structures. Clustering, improved minimisation protocols, and better criteria for sufficient convergence will be investigated to speed up this crucial stage of the process.

Together, these three approaches will lead to faster, more reliable and even more accessible DP4 analysis of NMR spectra. Original spectral data will be gathered from synthetic research groups in a range of universities, to test the robustness and utility of the process, and will increase the impact of the process. The quantitative analysis of confidence in conclusions drawn from NMR spectra will become more accessible, the quality of interpretation will increase, and the number of errors will be minimised.