ИСТИНА

PARAFAC is an attractive data exploration method in fluorescence spectroscopy because its mathematical formulation closely matches the underlying physical model for the fluorescence signal intensity in an excitation-emission matrix (EEM). Unfortunately, EEMs typically contain a scattering component, which doesn’t adhere to PARAFAC assumptions and therefore must be taken care of before the analysis. Interpolation of the wavelength region containing scattering signal [1] is typically used, owing to the low possibility of resulting local minima. This work takes a multitude of surface interpolation methods and compares their performance on model datasets consisting of simulated fluorescence and scattering signals. Compared performance metrics include root-mean-squared error of recovered intensity values and the Tucker’s congruence coefficient between loadings estimated by PARAFAC and their a priori values. The methods compared include monotone piecewise bicubic interpolation of each row or column of the EEM, multilevel B-splines, Whittaker interpolation with different combinations of difference orders, the LOESS algorithm and Kriging. The model datasets vary in correlation coefficients between columns of ground truth scores. The trends in recovery of the original fluorescence signal and in the PARAFAC estimation of the ground truth loadings are basically the same. The results (Figure 1) show that Whittaker interpolation, LOESS and Kriging provide the best results, subject to the need to cross-validate for optimal parameter values (Whittaker & LOESS) and high computational complexity (Kriging). The reported study was supported by the Russian Science Foundation (project number 21-77-10064). [1] Elcoroaristizabal S.; Bro R.; García J.A.; Alonso L.; PARAFAC models of fluorescence data with scattering: A comparative study; Chemometrics and Intelligent Laboratory Systems; 2015, 142, 124–130.