Unsupervised Machine Learning‐Based Clustering of Nanosized Fluorescent Extracellular Vesicles

Extracellular vesicles (EV) are biological nanoparticles that play an important role in cell-to-cell communication. The phenotypic profile of EV populations is a promising reporter disease, with direct clinical diagnostic relevance. Yet, robust methods for quantifying the biomarker content have been critically lacking, and require single-particle approach due to their inherent heterogeneous nature. Here, multicolor single-molecule burst analysis microscopy used detect multiple biomarkers present on single EV. authors classify recorded signals apply machine learning-based t-distributed stochastic neighbor embedding algorithm cluster resulting multidimensional data. As proof principle, use method assess both purity inflammatory status EV, compare cell culture plasma-derived isolated via different purification methods. This methodology then applied identify intercellular adhesion molecule-1 specific subgroups released by inflamed endothelial cells, prove apolipoprotein-a1 excellent marker typical lipoprotein contamination plasma. can be widely standard confocal microscopes, thereby allowing standardized quality assessment patient plasma preparations, profiling health disease.