BME PhD Preliminary Exam Announcement for Li Fang (F. Huang, advisor)

Everyone is invited to attend the public presentation beginning at 2:30pm.

Title: Molecular localization of single molecule switching sequences with deep learning

Date: December 12, 2022

Time: 2:30 pm

Location: MJIS 2001

Virtual link:

https://purdue-edu.zoom.us/j/99210793060?pwd=WnlHbVlSUlNGWE1KaGN4QWRCS0lEUT09

Meeting ID: 992 1079 3060

Passcode: 158594

Committee members: Dr. Fang Huang, Chair; Dr. Leopold N. Green; Dr. Young L. Kim; Dr. Chongli Yuan

Abstract: Single molecule localization microscopy (SMLM) has become an essential tool in imaging nanoscale biological structures. It breaks the diffraction limit by utilizing photo-switchable or photo-convertible fluorophores to obtain isolated single molecule emission patterns (i.e. PSFs) and subsequently localize the molecule’s position with a precision down to ~ 25 to 80 nm laterally-axially. However, standard SMLM algorithms require sparse activation to minimize emission pattern overlapping, which limits imaging speed and temporal resolution and hinders its application in dynamic live cell imaging. Some multi-emitter fitting algorithms have been developed to analyze higher density data and some methods have shown improved localization precision when utilizing multiple frames. But these methods require an accurate PSF model and prior knowledge of photo-switching behavior and fail to fully extract the information in the sequence data. Here, we propose to develop a high-density single molecule localization algorithm through molecule blinking sequence analysis with deep learning to achieve high localization precision approaching the theoretical limit at high density in three dimensions. We aim to apply the algorithm to resolve nanoscale intracellular structures and fast dynamic processes in live cells with high temporal and spatial resolution.