Wednesday, June 10, 2009

Bernhard Jaun

Q. What's your position and where are you working?
Professor in Organic Chemistry at ETH Zurich (Swiss Federal Institute of Technology). Head of the NMR labs.

Q. Where have you been working before?
Columbia University New York, then ETH Zurich for the last 29 years.

Q. Briefly describe your research.
Physical organic chemistry, in particular its application to biological questions.

Q. What do you use NMR for?
I am the head of NMR operations in an institute with more than 200 scientists. Most of them use NMR. The applications go from 3D solution structures of biopolymers to physical organic applications of NMR (such as host guest complexation, dynamic processes, thermodynamics and kinetics) to structure elucidation of novel natural compounds and (for the majority) characterization of synthetic intermediates.

Q. Which NMR software are you using now?
Topspin, VNMR, Mnova, iNMR, plus specialized software for 3D solution structure calculation such as XPLOR, SPARKY, CNS, DYANA, MARDIGRAS etc.
We have quite a large percentage of people using Macs (ca. 50%) in our institute.

Q. Which other NMR software have you used in the past?
SwaNMR for Mac OS9 and most of the software used for NMR over the last thirty years.

Q. How do you rate iNMR?
iNMR is a good and fast program which can do practically all of the work an NMR spectroscopist will ever need. Its strength lies in the flexibility and its more mathematical/physical approach to NMR such as beeing able to do all kinds of transforms, simulating dynamic exchange problems, analyzing spin systems etc. Clearly, the program is written by an NMR specialist for NMR spectroscopists.

The "weaknesses" are in the fact that iNMR is not as easy to learn as some other programs by people who do not know much about NMR and are only interested in getting "nice" plots and listings for their synthetic papers etc. Compared to other programs, iNMR uses only a fraction of icons and palettes but insiders can work very efficiently because of all the keyboard shortcuts and the scripting ability. The current versions still have a few bugs or inconveniences in the field of graphics, e.g. when it comes to plot 1D spectra at the border of 2Ds etc., axis adjustments when changing the window size etc. [Editor's note: this interview refers to the old version 3; the current version 4, made with the collaboration of prof. Jaun himself, solved all the above mentioned problems].

Maybe the best point about iNMR is that according to my experience, there is no other software where the programmer is so fast in responding to either bug reports or demands for new features. So, if I still sometimes get angry about a bug (or something I want to do but cant find out how) in iNMR, it is usually my own fault because I did'nt take the time to write to the author about it. If I had contacted the author, the problem would long be solved by now. Compare that to MS Office or the spectrometer manufacturers NMR programs!

Q. Is it enough for your needs?
We NMR spectroscopists have to accept that for a majority of the scientists in todays chemistry/biology research, NMR is a black box that's neverless - and "unfortunately"- absolutely necessary. They like to use software that seems to generate listings and plots without requiring knowledge by the operator. We still try to teach our own students about the innards of NMR-experiments. But the reality is that black-box attitude and the trend for automation are increasing all the time.

I think that in my domain of responsability with 200 scientists using NMR, it might actually be a good idea to start to write some scripts for iNMR that do all the standard processing for routine spectra. This might make iNMR more poular for all those, who are not really interested in the inner workings and just need a nice plot to show to their supervisor and who now rather use MNOVA for Mac because they think it is easier to use.

Also, I think that iNMR could become the tool of choice for all special and more physical things that can be done by NMR. In particular, there is only a very limited number of still living programs that can iteratively fit dynamic spectra from complicated exchanging spin systems. Other things I could think of are extracting coupling constants from 2Ds by simulation of cross peaks, analysing relaxation data, measuring residual dipolar couplings from heteronuclear 2D spectra, diffusion etc. etc.


At 5:45 PM, Blogger old swan said...

This is one of the finest posts on my own blog. As a form of respect, I prefer to reverse the roles: the words of the guest are under the spotlight, mine are here amid the foot notes. Just a notation: in theory, if NMR is to work like a black-box, there is no space for third-party software: everything could be done on the spectrometer and that's all. In practice the future is third party software that acts as a black box, like prof. Jaun hints. Huge amounts of plots, peak listings and the likes will be produced and stored. Another computer will read them and the role of the chemist will be redefined: he will become a middle-man among many computers. Anyway, observation is a fundamental skill, today at least.


Post a Comment

<< Home