Bioinformatics and cheminformatics are established disciplines, but nanoinformatics, the development of computational tools for understanding and designing nanomaterials, is still in its infancy. In light of the new data-driven approaches for nanomaterials discovery, there is a growing need for in silico tools tailored to analyze nanomaterials data sets. This is particularly crucial for materials where a crystalline structure cannot be obtained, and therefore, the characterization data sets are less structured, and there are no standard methods for data mining. Here, we present a computational framework capable of visualizing, describing, and comparing nanoparticle data sets obtained with super-resolution microscopy at the single-particle and single-molecule level. Our method allows us to: (i) visualize multiparametric nanoparticle data sets to grasp material properties and heterogeneity; (ii) have a quantitative evaluation of a material through a series of molecular descriptors, and (iii) compare different materials quantitatively and globally, going beyond comparison of a single property. We applied this method to a library of targeted nanoparticles revealing particle heterogeneity, similarities, and correlations between the synthesis and the physicochemical properties of the different nanomaterials. Finally, we show the potential of this approach to reveal batch-to-batch variations in time and between users hidden in standard analysis.
Izquierdo-Lozano et al. (Wed,) studied this question.