Figure 2.

Graphical representation of the physical model formulating a smFRET experiment. Fluorophores and molecule $f_n^{\text{D}}$, $f_n^{\text{A}}$, and s_n evolve stochastically through time (left to right). Measured photon intensities $I_n^{\text{D}}$ and $I_n^{\text{A}}$ in the donor and acceptor channels are determined by (i) the photoemission rates associated with s_n, (ii) the photophysical state of the fluorophores, and (iii) shot noise. FRET efficiency ${ϵ}_{s_n}$ depends exclusively on the molecule state s_n. By contrast, apparent FRET efficiency ${ϵ}_n^{*}$ depends on the measured intensities and, due to photoartifacts, need not coincide with ${ϵ}_{s_n}$. Following common convention, in the schematic, stochastic variables are denoted with circles, deterministic variables are denoted with boxes, measurements are shown shaded, and dependencies among the various variables are indicated by arrows. Dashed lines indicate dependences irrelevant to the generation of the measurements.