Retrospective multivariate analysis of pharmaceutical preparations using H nuclear magnetic resonance (NMR) spectroscopy: Example of 990 heparin samples.

NMR spectroscopy has become an essential diagnostic tool for pharmaceutical preparations worldwide, with many runs performed annually on a routine basis. In this study multivariate analysis was performed on a big dataset of 990 NMR heparin spectra recorded over six years (2012-2017) in our laboratory. Several steps of statistical analysis of accumulated data were used to differentiate samples according to animal origin (bovine, porcine and ovine heparin), purity grade (crude and purified heparin), distributing company as well as to estimate their closeness to the heparin reference sample (SST) provided by US Pharmacopeia. The projection of new samples on these models can automatically forecast of all mentioned qualitative heparin properties within one minute. NMR data of heparin pharmaceutical preparations provided a good opportunity to demonstrate the potential of exploratory data analysis considering the same sample preparation and acquisition parameters were used.