Tuesday, March 4, 2008

Basic Misinterpretations of NMR-Data

I have created a series of pages on my webserver dealing with misinterpretations, typos and any other type of errors within NMR-data. I am definitely not talking about errors below 10ppm - I am talking about errors, which can be easily detected by application of appropriate computer algorithms using a few seconds of CPU-time.

At the moment 2 examples are online - I promise 'More to come' ! - Stay tuned, check back !

http://nmrpredict.orc.univie.ac.at/csearchlite/NMR_misinterpretation.html

In order to do a serious job I have to cite every paper in error I find during my daily work - BUT I dont want to blame somebody personally. On the other hand I think its necessary to analyze the quality of available NMR-data, because this is the basis for solving future structure elucidation problems ! Keep in mind, what is necessary to perform this task: State-of-the-art algorithms for automatic data-checking with an underlying database of highly verified spectra AND the largest CNMR-database available (despite its size of more than half a million C-spectra it is still incomplete)

5 comments:

Anonymous said...

Already 4 examples online

ChemSpiderMan said...

Great Stuff Wolfgang...did you see my post on Obtusallenes?

http://www.chemspider.com/blog/adding-publication-details-to-chemical-structures-on-chemspider.html

Any idea where you would have stood on that? Did you catch the error before adding to the database by any chance?

Wolfgang Robien said...

There are already 6 examples online - the last one uses HH-COSY, HMQC and HMBC

hko said...

Comment to NMRShiftDB - "Latest addition" using HH-COSY, HMQC and HMBC

Is this an ambiguous and obscure case of misinterpretation
of nmr data ? On the one hand we have qualified shift data from
experiments like cosy, hmqc and hmbc, and on the other hand we
have values from prediction tools.
Whom we should trust, the experiment or the prediction ?

The answer should be given in this case by the reviewer of NMRShiftDB,
provided that he makes the job with thoroughness and experience.
Depending on his knowledge he has to decide whether the assignments
are reliable or dubious.
Since there are many well assigned shifts of indole derivatives in the
NMRShiftDB database, the decision should be easy and obvious,
as demonstrated by Wolfgang, who detected at least eight clear
misassignments.
Conclusion:
The bad message is that the reviewer did the job not properly.
The good message is that the data are open and may be curated easily.
Therefore many thanks to Wolfgang to make the errors public thus
enabling quick error detection and correction.

Wolfgang Robien said...

ad hko:

Thanks for your detailed comment. I agree, that curation is easy because the data are open. The necessary step BEFORE curation is detection of errors. This detection has been done using CSEARCH-algorithms, which have been implemented into NMRPredict.

For me personally the main problem remains: There is an OPEN system, which influences many 1000s of chemists during their daily work of CNMR-assignments, which has no state-of-the-art (or at least rudimentary) data checking protocols. This particular entry has a rating of '10' (highest), despite there are severe deviations between experimental and predicted chemical shift values. The error detection on the NMRShiftDB-dataset has been done by the more 'professional' versions of prediction programs named ACD/CNMR-Predictor, CSEARCH and NMRPredict.