iNMR news #69

[GB, author of iNMR]

Today I am going to introduce an exciting new iNMR module.
Let me recapitulate first the performance of the latest two versions.
Version 3.2.2 of iNMR was almost perfect, apparently bug-free. After
the arrival of OS 10.6 (Snow Leopard), it look that the right moment
for an accurate test had also arrived.
Version 3.2.2 was therefore tested more than any other version of the
present year and proved to be really the best one. Three memory leaks
were found, but none of them was a real issue. Nonetheless, a new
corrected version (3.3) was compiled and published, in which all the
known problems had been removed. This happened one month ago and no
bad news have arrived so far. I conclude therefore that the most
reliable version of the year is actually version 3.3. I urge you to
upgrade, unless you are already using version 3.2.2, because some
older version (like 3.1.5) are certainly less reliable.

http://www.inmr.net/downloads/iNMR_3_3.dmg

Progress goes on, however: today's version is numbered 3.3.1 and
contains a novel module to simulate spectra. It's the fourth module in
the series and they are all necessary. Let me summarize:
- The J Manager calculates the Js and the shift of a first-order
multiplet (ideal case);
- The spin system simulator takes into account both the quantic
effects and the chemical exchange, but assumes that the line-shapes
are accurately described; in other words, it is limited to 1-D
spectra;
- The deconvolution module offers a different approach, but suffers
from the same limitations.
Summarizing, up to now there was nothing to simulate 2-D spectra or
similar cases, where the number of points is very very low and/or the
shape of the peaks cannot be described with a simple mathematical
function. A typical case where such a tool could be precious is when
you want to measure the coupling constants directly from a 2-D
spectrum, because you have a big molecule (like a natural product) and
some signal cannot be distinguished in a 1-D spectrum (think at the 1-
D spectra of glucose or saccharose).
These kind of problems can now be handled by the "Fourier Pairs
Simulator". From the user's point of view, it works on a fragment (of
a section of) a 2-D spectrum. Internally, instead, it simulates the
time domain spectrum (FID), applies the same weighting functions that
have been applied to the experimental spectrum and compares the two.
The accuracy of these simulations can be really astonishing, and no
special skill is required: the new module is actually easier to use
than the older companions. You will probably use it for 1-D spectra
too.
If you are interested, and you really should, please repeat the
tutorial:

http://www.inmr.net/articles/DQFJ.html

or jump directly to the manual:

http://www.inmr.net/Help3/task/jsimul.html

Just remember that it's a completely new tool and needs some testing;
it will likely improve in the next few months.
I really don't know if something similar has already been invented. It
is based on well-known theory. Simply it's more convenient to
implement such a tool into the same program that performs the Fourier
Transform, because this is the simplest way to ensure that the exactly
same operations are performed on the experimental signal and on the
simulated signal.
There is always space to invent something new. If you have an idea and
don't know how to realize it, I am here to collaborate. If you simply
need assistance in using the new simulator, I am still here to help
you.