We are using an old version of UXNMR (920801) and have noticed what may
be a problem with peak integrations on our AMX-400. The background to
this is that we acquired several series of sequential control spectra on
a standard protein to test integral variability amongst the files on a
well resolved peak. We processed the data files with identical
processing parameters in a macro, use the AI-cal feature, and always
integrate the same ppm region to reduce operator bias. In UXNMR the
integral standard deviation was about 15%, with variations of up to 40%.
This had us suspecting a hardware instability until we noticed by visual
inspection of hard copy that the integral variation seemed far in excess
of the variation in peak area in the plots. Exporting the data to Felix
and integrating gave a standard deviation of 3%, with a largest
variation of 9%. To make a long story short, is anyone aware of a
problem/bug with integrations in UXNMR in general or version 920801
specifically?
Please email me any responses and I can post a summary if needed.
Jeff de Ropp
NMR Facility, UC Davis
jsderopp@ucdavis.edu
916-752-7677
916-752-3516 Fax
NMR Facility Web site URL http://www.nmr.ucdavis.edu