Graph Representation Learning for Structural Variation Discovery
Name
cui-tacui-meng-eecs-2026-thesis.pdf
Size
4.28 MB
Format
Adobe PDF
Checksum (MD5)
11600c4dea0340dacb2ff021af8c6347
Author(s)
Cui, Tony Ao
Advisor(s)
Berger, Bonnie
Popic, Victoria
Date Issued
February 2026
Publisher
Massachusetts Institute of Technology
Abstract
Structural variants (SVs) are large-scale genomic alterations and key drivers of genetic diversity in the genome, often implicated in a wide range of genetic diseases. Despite their significance, our understanding of the distribution and diversity of structural variation in the genome remains limited due to the considerable diversity of SV classes and the complexity of their signatures. Though there’s been extensive effort in the development of SV callers, which detect and classify SVs into predefined class sets, there has been limited work on the discovery of novel classes of SVs. This paper introduces NeuSV, the first graph representation learning framework for discovering novel classes of complex structural variants.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
