Independent component analysis disentangles fluorescence signals from diffusing single molecules

Multiparameter single-molecule fluorescence measurement is a promising technique for detecting and quantifying heterogeneity in freely diffusing molecules, albeit with challenges in data analysis due to limited photon numbers. Here, we present a universal analytical framework that leverages independent component analysis for unmixing multiparameter fluorescence signals. Applications to static and dynamic mixture systems demonstrate its potential allowing model-free separation of subpopulations with microsecond time resolution in nanomolar concentration regime.