Enhancing Fluorescence Correlation Spectroscopy with Machine Learning for Advanced Analysis of Anomalous Diffusion

TL;DR

This paper introduces a machine learning-based method to enhance Fluorescence Correlation Spectroscopy (FCS), enabling more accurate analysis of anomalous diffusion in living cells with reduced limitations and improved capabilities.

Contribution

The paper presents a novel machine learning approach that extends FCS analysis to a broader range of anomalous motions and reduces acquisition time constraints.

Findings

Enlarges the range of detectable anomalous diffusion behaviors.

Validates the method with experimental data on fluorescent beads.

Achieves analysis performance comparable to advanced single-particle-tracking algorithms.

Abstract

The random motion of molecules in living cells has consistently been reported to deviate from standard Brownian motion, a behavior coined as ``anomalous diffusion''. Fluorescence Correlation Spectroscopy (FCS) is a powerful method to quantify molecular motions in living cells but its application is limited to a subset of random motions and to long acquisition times. Here, we propose a new analysis approach that frees FCS of these limitations by using machine learning to infer the underlying model of motion and estimate the motion parameters. Using simulated FCS recordings, we show that this approach enlarges the range of anomalous motions available in FCS. We further validate our approach via experimental FCS recordings of calibrated fluorescent beads in increasing concentrations of glycerol in water. Taken together, our approach significantly augments the analysis power of FCS to capacities that are similar to the best-in-class state-of-the-art algorithms for single-particle-tracking experiments.