M.S. Defense: Bill Manka

Please attend:

Master’s Thesis Oral Defense

Title: Unsupervised detection of G-Quadruplex DNA structures using nanopore current signatures

M.S. Candidate: William (Bill) Manka

Major Advisor: Dr. Derek Aguiar

Associate Advisor: Dr. Ion Mandoiu, Dr. Yufeng Wang

Date/time: Wednesday, August 9, 2023, 12:30pm

Location: ITE 201 or Webex

Meeting link: https://uconn-cmr.webex.com/uconn-cmr/j.php?MTID=mbf8097bc1b09b4d239b0ad0c798b1f87

Meeting number: 2622 767 9893

Password: uV4PP3RRVd9

Abstract:
G-Quadruplexes (G4s) are prominent non-canonical DNA structures that play important functional roles including the regulation of gene expression. Nanopore sequencing is a third generation DNA sequencing technology that produces long sequencing reads by unwinding the DNA double helix and passing it through a biological pore. An electrical current that passes through the pore is disrupted by the DNA occupying the pore, which enables the prediction of DNA bases. We hypothesize that, since G4 structures are relatively large and complex compared to the chain of single bases passing through the nanopore, their arrival at the pore might cause characteristic disruptions in the flow of current across the pore channel. First, we compare current signatures centered around G4 motifs to canonical DNA across the human genome. Based on this data analysis, we formulate the problem of G4 prediction from nanopore current as unsupervised clustering of time series data. We identify a characteristic trace signature among G4 motif regions that not only allows for highly separable clusters, but also supports the hypothesis above, suggesting that G4 structures can be detected by nanopore current occlusions.