Tucker3 modelling of EEMF spectroscopic data sets using restrained initialization of spectral variables: fluorimetric analysis of mixtures consisting of bioactive molecules. (14th April 2021)

• Record Type: Journal Article
• Authors: Kumar, Keshav
• Is Part Of: Luminescence. Volume 36:Number 5(2021)
• Page Start: 1265
• Page End: 1271
• Publication Date: 2021-04-14
• Subjects: biomolecules -- convergence -- EEMF -- modelling -- Tucker3 -- variable initialization; Luminescence -- Periodicals; Bioluminescence -- Periodicals; Chemiluminescence -- Periodicals; Luminescence -- Periodicals; 535.35
• DOI: 10.1002/bio.4052
• Languages: English
• ISSNs: 1522-7235

Abstract

Abstract: The excitation−emission matrix fluorescence (EEMF) spectroscopic technique provides a viable means of analyzing samples from different fields. EEMF spectral data sets are much larger in volume, so that they can only be interpreted using novel data analysis techniques. Here, a novel spectral initialization approach was introduced to fit the Tucker3 model to EEMF spectral data. The suggested method involved variable initialization in a restrained way, yielding initial estimates of EEMF spectra that were comparable with experimentally acquired EEMF profiles. Tucker3 modelling of EEMF spectra with these initial estimates made these analyses fast and computationally economical. The Tucker3 model with the proposed initialization approach was found to yield much purer spectral and concentration profiles. The proposed approach was validated by successfully processing the EEMF spectral data sets of biomolecule mixtures. Abstract : The present work introduces a novel spectral initialization approach to fit the Tucker3 model to EEMF spectroscopic data sets of dilute aqueous mixture of bioactive molecules.