Novel Data Evaluation Algorithms: J-CorrelatorTM
aCarlos Cobas, aPablo Monje, aSantiago Fraga, and bStanislav Sykora
aMestrelab Research, Santiago de Compostela, Spain
bExtra Byte, Castano Primo, Italy
presented at 49th ENC Conference, Asilomar, CA (USA), March 9-14, 2008.
We have developed a family of fast algorithms  which opens new avenues in NMR data evaluation. This poster illustrates the results obtained with what we call the J-CorrelatorTM - a member of the family which permits to carry out fast, visual correlation between various spectral-line splittings in complex 1D NMR spectra.
The J-CorrelatorTM operates on phased, real part of a 1D NMR spectrum and generates a 2D plot in which the additional (vertical) dimension encodes a splitting value. In this approach, an isolated single line would not appear anywhere in the 2D plot. Any pair of lines, however, gives rise to a cross-peak located at the center of the pair along the horizontal dimension (ppm) and at the splitting value (Hz) along the vertical dimension. Distinct cross peaks which share the same splitting value and thus lie on the same horizontal line indicate a probable J-coupling between the multiplets to which they belong. Vice versa, a cross peak which has no partner with the same splitting value is spurious and indicates a probable absence of a coupling. The advantage of this for an expert attempting to interpret the spectrum in terms of a molecular structure is evident.
J-CorrelatorTM is quite noise-tolerant and generates very clean peaks. Preliminary results indicate that the same type of algorithms is applicable also to the recognition of splitting patterns other than doublets (triplets, quadruplets, and combinations thereof).
Another fascinating possibility is the combination of J-CorrelatorTM with the 1D Resolution BoosterTM [1,2,3]. This sharpens the peaks even further and, due to the resolution of overlapped multiplets, enhances the power of J-CorrelatorTM as a novel MSCA (Multiplet Spin Coupling Analysis) tool .
The J-CorrelatorTM, complete with an algorithm for automatic removal of spurious splitting cross-peaks, has been implemented and is being tested on practical data. Further work on its optimization and generalization is in progress, but its present performance makes it already invaluable in molecular structure verification and/or elucidation.
J-CorrelatorTM references and links:
- Sykora S., Cobas C., to be published.
- Cobas C., Larin N., Iglesias I., Seoane F., Sykora S.,
Novel Data Evaluation Algorithms: The 1D and 2D Resolution BoosterTM,
poster submitted for the 49th ENC Conference, Asilomar, CA (USA), April 9-14, 2008.
- Selective Resolution Booster by Cobas C. on his blog.
- Sykora S., Cobas C., Advances in Computer-Assisted Evaluation of NMR Spectra,
presentation at SMASH 2007, Chamonix, France, September 16-19, 2007.