In case you didn’t know, version 11 of ACD/Labs software was released in November of 2007.
While there are several updates to the software over the course of a year, a major new version of each of our NMR products is released on an annual basis.
Over the next couple of weeks on this blog, I will try to mix in some new features and improvements that have been implemented in version 11.
First up, I’ll talk a bit about improvements in the automated multiplet analysis algorithm in ACD/1D NMR Assistant, ACD/1D and 2D NMR Processor, ACD/1D and 2D NMR Manager, and ACD/1D and 2D NMR Expert.
For those unfamiliar with this process, automated multiplet analysis refer to the use of a software algorithm that automatically characterizes coupling patterns and extracts coupling constants from multiplets in an NMR spectrum.
The most popular application of this feature is that it provides an incredibly fast way of generating a formatted multiplet report for patents and publications:
1H NMR (400 MHz, DMSO-d6) d ppm 2.34 (dd, J=15.97,
8.06 Hz, 1 H) 2.64 (dd, J=16.05, 5.35 Hz, 1 H) 3.81 (tt, J=7.71,
5.33 Hz, 1 H) 4.47 (d, J=7.47 Hz, 1 H) 4.87 (d, J=5.13 Hz, 1 H)
5.68 (d, J=2.20 Hz, 1 H) 5.88 (d, J=2.34 Hz, 1 H) 6.58 (dd, J=8.13,
1.98 Hz, 1 H) 6.68 (d, J=8.06 Hz, 1 H) 6.71 (d, J=1.91 Hz, 1 H)
So how well does this automated algorithm perform in Version 11?
Significant development efforts were put into improving the automated routine for Version 11 and I will present the results of a direct comparison to version 10 of the software.
For this study, two different data sets each consisting of the 1H NMR spectra of 30 samples were studied (~250 multiplets in total).
Test Set 1- A set of 30 spectra (~250 multiplets) with reasonably good signal to noise:
In version 10, 53% of the multiplet patterns in the spectra for the 30 samples were correctly defined, 32% were undefined, leaving 15% that were incorrectly defined. Multiplets that are termed as undefined are simply given a "m" designation for multiplets. In standard practices and manual analysis the "m" designation is often assigned by users to multiplets with unresolved peaks, often due to strong coupling.
Version 11 Results:
As you can see, while the number of correctly defined multiplets increased by 5%, the most notable observation is the reduction of incorrectly defined multiplets from 15% to 1%.
Test Set 2- A set of 30 spectra (~250 multiplets) with lower signal to noise:
Version 10 Results:
Version 11 Results:
As you can see in the second study, we did sacrifice some correctly defined multiplets (6%), but we were able to reduce the number of incorrectly defined multiplets significantly from 14% to 3%.
So how does this improved accuracy impact the end user?
First of all, this will result in significant time savings with more accurate automated multiplet analysis and report creation for spectroscopists’ and chemists’ patents, publications, etc.
But perhaps of more scientific relevance, is that improved automated multiplet analysis heavily impacts both the performance of automated structure verification in ACD/1D and 2D NMR Expert, as well as the structure verification algorithm included in ACD/1D NMR Assistant.
How much impact are we talking about here?
Stay tuned, it will be the topic of my next post.
If you are currently using Version 11, try it out, and share your results in the comments section.