With the rapid advancement of fluorescence microscopy, there is a growing interest in the multiplexed detection and identification of various bioanalytes (e.g., nucleic acids and proteins) for efficient sample processing and analysis. We introduce in this work a simple and robust method to provide combinations for micrometer-scale fluorescent DNA barcodes of hierarchically assembled DNA origami superstructures for multiplexed molecular probing. In addition to optically resolvable dots, we placed fluorescent loci on adjacent origami within the diffraction limit of each other, rendering them as unresolvable bars of measurable lengths. We created a basic set of barcodes and trained a machine learning algorithm to process and identify individual barcodes from raw images with high accuracy. Moreover, we demonstrated that the number of combinations can be increased exponentially by generating longer barcodes, by controlling the number of incorporated fluorophores to create multiple levels of fluorescence intensity, and by employing super-resolution imaging. To showcase the readiness of the barcodes for applications, we used our barcodes to capture and identify target nucleic acid sequences and for simultaneous multiplexed characterization of binding kinetics of several orthogonal complementary nucleic acids.
Keywords: DNA barcodes; DNA origami; machine learning; multiplexed probing; self-assembly.