SwaN-MR from Infancy to Maturity

Giuseppe BalaccoMenarini Ricerche S.p.A., Via Sette Santi 3, I-50131 Firenze, ItalyThe NMR software SwaN-MR was described in the scientific literature more thanfive years ago. During this period the program has met worldwide success, and hasbeen modified many times. The present article corrects and updates the materialpublished in 1994. The task of processing and interpreting NMR spectra will bediscussed in general terms. The unique approach of SwaN-MR is stressed. It is centredon the interaction between scientist and spectrum. As an example, the techniquesarising from manual phase and baseline correction have been extended to the caseof fitting an experimental spectrum with a simulated one.IntroductionNuclear Magnetic Resonance, especially at higher magnetic fields, has the advantage,compared to other spectroscopies, that every single signal in the spectrum is easilyinterpreted or, in other words, assigned to a given nucleus. Given that an n-dimensionalspectrum can contain hundreds of signals, the amount of information that can be extractedis substantial. Normal practice suggests that this information must be extracted completelyin order to be sure. Any interpretation of a spectrum unable to explain all of its features isprobably erroneous. This process of extraction has always been performed by a humanbrain and can take from a couple of minutes up to months, depending on the complexityof the problem. Although many efforts have been spent teaching computers to performthe task, we cannot say when this will happen. It seems that spectroscopists are quitehappy performing the interpretation because it is a challenging and rewarding intellectualactivity. In the old days of continuous wave NMR, the spectrum was plotted on a piece ofpaper during acquisition and studied away from the spectrometer, on the researcher’sdesk. With the advent of computers and FT-NMR, a second opportunity arose: to studythe spectrum interactively on a computer screen. Although this practice began directly onthe spectrometer, it is clear that the best place is still the desk. In 1994 the author reportedsome problems connected with this desktop analysis and proposed his solution in a paperwhich, after explaining how to transfer spectra onto a Macintosh computer, described thenew program SwaN-MR (Balacco, 1994). It is well known that in the six years since thenthere has been a constant evolution in the world of computers. It is arguable that somethingchanged in NMR software too. SwaN-MR also underwent a constant evolution. Everypart of the program has been rewritten at least once, many new functions have beenintroduced, others were eliminated. In many aspects (speed, stability, compactness) SwaN-MR appears to have reached the top. The program is also much more productive and