A numerical method for detecting and classifying multi-channel and multi-state binding events in single-molecule TIRFM experiments

Biomolecular binding may be observed using Total Internal Reflection Fluorescence Microscopy. By using multiple channels, which corresponds to different colors of fluorescence, the real-time interaction of multiple proteins can be observed. Although for many of these interactions the bound state is binary, for some the binding is more complex, with discrete states demonstrating intermediate levels of fluorescence. We will explain a numerical method for extracting these fluorescence traces from microscope data and then organizing and classifying traces from single-molecule experiments. With multiple channels and multiple binding states possible, the possible binding patterns are numerous. The data is organized in groups by these binding patterns and the statistics of each interaction is displayed. Which combinations are most likely, the binding duration of each protein or the probable binding/dissociation order of each subunit are all data which can inform on the underlying mechanism behind the interaction. We show how our method determines cooperative or inhibitory binding in multi-protein systems, or observe the construction of protein complexes.