MBioHOS |
This advanced plugin needs a specific license (on top of licenses for the NMR and Chemometrics modules).
Mnova BioHOS is designed for the NMR analysis of biotherapeutics, like monoclonal antibodies. 2D or stacked NMR spectra of reference samples are compared with that of each test sample using a fingerprinting, or spectral similarity determination.
Includes four published and accepted methods to conduct analysis, each with its set of tailor-made visualizations allowing you to drill in results:
Combined Chemical Shift Difference, CCSD – number that represents the shifts in peak positions
Easy Comparability of Higher Order Structure, ECHOS – simple representation of spectral differences
Principal Component Analysis, PCA – unsupervised chemometrics
The 1D PROFILE experiment consists in measuring the correlation between pairs of spectra by comparing their: •contours: the broadened spectrum •fingerprints: the result of subtracting the contour to the spectrum
The results of the PROFILE experiment are a table of correlations and a diagram of percentiles. These percentiles show the statistics of the intra-class correlations and inter-class correlations. To identify the class of each spectrum it is used the Class value, which can be edit in the Stacked Items table. Therefore, it is very important to classify the input spectra before running PROFILE. The color code used in correlations table and percentiles diagram: •only color: intra-class correlation •two colors: inter-class correlation •grey color: no-class defined The resulting correlations and percentiles are shown after running the experiment. Correlations are normalized in a logaritmic scale, so the same pair of spectra should have the correlation value = 100, the maximum possible value. The higher the correlation, the more similarity between spectra. When the percentile of an inter-class is in similar ranges than its corresponding intra-classes, it means that the spectra of both classes are "similar". You will find further information and resources about this plugin, here and in this application note. |