Initial Guesses Generation for Fluorescence Intensity Distribution Analysis

Abstract

The growing number of applications of Fluorescence Intensity Distribution Analysis (FIDA) demands for new approaches in data processing, which should increase speed and robustness of the method, iterative algorithms of parameter estimation, although proven to be universal and accurate, require some initial guesses (IG) for the unknown parameters. To develop a general algorithm for IG generation is not a trivial task, but in case of FIDA, we are in an advantageous position since available ranges of parameters values are restricted due to numerical limitation of the used model.

In the present work we introduce an approach for IG generation in FIDA experiments (sample parameters - brightness and concentrations, and an experimental set-up related parameters - background, observation volume profile) based on the method of moments. A number of analytical simplifications have been introduced in order to increase the accuracy and robustness of the numerical algorithms. The performance of the developed method has been tested on simulated and experimental data. Iterative fitting with calculated IG proved to be at least five times (and often more) faster than with the arbitrarily chosen IG. Using simulated data we have also shown that at least three sets of parameters a and b, characterizing the observation volume, ensure a comparable quality of fit. A rule for selection the best set is suggested. An outline of further investigations to explore comprehensively statistical robustness of the parameters is given.