Abstract
An integrated approach based on the use of data mining methods has been proposed to improve the efficiency of the analysis of photon counting histograms in the study of the molecular composition of a substance by the method of fluorescence fluctuation spectroscopy. The method of principal components is used to test the hypothesis about the cluster separability of multidimensional experimental data. The reason for the compression of a point cloud into a characteristic nonlinearity, or so-called arc-shaped cloud, in the space of first two principal components is investigated. Examples of simulated data sets on some selected molecular systems of various brightness and concentration are considered. Nonlinear effects complicate interpretation and subsequent quantitative analysis of data. It has been established that the arching of the data cloud is a consequence of the presence of a significant variation in one or more physical parameters. In particular, it is the result of a significant increase in the variation in the parameters of the brightness or concentration of molecules. These parameters can be as additional measure in assessing the quality of the experiments if only one type of molecule is studied, and also can be used for characterizing the system under study in the case of a mixture of molecules of different types. It is proposed to apply the locally weighted scatterplot smoothing normalization to eliminate the nonlinear effects in the space of principal components.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call