An information-theoretic framework for optimizing FRET microscopy imaging system

Ivan Sigmund^1*, Keita Kamino¹

¹Institute of Molecular Biology, Academia Sinica, Taipei city, Taiwan

* Presenter:Ivan Sigmund, email:ivansigmund@as.edu.tw

Three-cube Förster Resonance Energy Transfer (FRET) microscopy is a powerful technique used to infer the degree of molecular interaction in various biological systems, including chemotactic signaling pathways in E. coli. However, selecting the ideal elements of the optical system for a given fluorophore pair presents a non-trivial challenge, because each element involves trade-offs. For example, selecting emission filters with wider bandwidths increases photon collection but also crosstalk between channels. Yet, a framework for systematic, quantitative comparison of alternative imaging system configurations is still lacking. Here, we present an information-theoretic approach based on mutual information that enables such comparisons for a given fluorophore pair and imaging system. We demonstrate the usefulness of the framework by finding the optimal emission path filters, dichroic mirrors, and even a camera system for a reference FRET pair and excitation configuration. This framework, along with our open-source code, enables researchers to computationally evaluate alternative filter configurations using a single experimentally calibrated reference measurement, thereby avoiding the need to empirically test each potential setup.

Keywords: Information-theoretic optimization, Optical system design, Mutual information, FRET microscopy