ORCID ID

https://orcid.org/0009-0001-2256-8823

Date Awarded

2023

Document Type

Thesis

Degree Name

Master of Science (M.Sc.)

Department

Chemistry

Advisor

Kristin Wustholz

Committee Member

Nathan Kidwell

Committee Member

Christopher Abelt

Abstract

Visualizing nanoscale biological systems allows us to uncover their detailed structure and functions, which have major implications in the biomedical field. Super-resolution microscopy is a powerful tool for fluorescence imaging because, by overcoming the diffraction limit of light, it accesses structural detail with unprecedented spatial resolution. Although multicolor super-resolution imaging has been successfully implemented in many experiments, its efficiency is limited by reliance on spectral measurement for emitter identification, which limits the combinations of compatible probes to be used together. Blinking-based multiplexing (BBM) is a novel approach that circumvents the need for spectrally-distinct emitters by instead exploiting the intrinsic differences in their blinking dynamics, or the stochastic fluctuations in emissive and nonemissive intensities of single-molecules under continuous photoexcitation. We find that BBM is most efficiently carried out using multinomial logistic regression (LR) to classify hundreds blinking dynamics obtained through single-molecule spectroscopy (SMS). Blinking dynamics are captured for three emitters—quantum dots (QD), rhodamine 6G (R6G), and pyrromethene 605 (PM605)— both on glass substrate and in complex poly(vinyl alcohol) (PVA) matrix for analysis with LR. Our results show that LR rapidly generates highly accurate predictive models for a variety of emitter systems under many experimental conditions.

DOI

https://dx.doi.org/10.21220/s2-sam6-a089

Rights

© The Author

Available for download on Sunday, August 25, 2024

Included in

Chemistry Commons

Share

COinS