Exploring Structural Variation in Tumor Evolution through Nanopore Sequencing

Structural variants (SVs) are genomic alterations encompassing deletions, insertions, and segment rearrangements, ranging from kilobases to entire chromosomes. Despite their significance as biomarkers in oncological diseases, these variants have remained relatively unexplored compared to single nucleotide variants, largely due to the inherent limitations of short-read sequencing technologies that have dominated large-scale genome sequencing projects. This scenario has undergone a transformative change with the advent of long-read sequencing technologies, which have enabled the achievement of the first truly complete human telomere-to-telomere reference genome, successfully filling gaps that short reads could not resolve. This project focuses on conducting a comprehensive performance evaluation of long-read-based structural variant callers, specifically in the context of tumor evolution analysis. To address the limited availability of appropriate datasets, we have developed specialized workflows leveraging high-performance computing resources for generating synthetic data with custom SVs, thus facilitating robust benchmarking of various structural variant detection methods. This computational approach enables systematic evaluation of SV detection algorithms under controlled conditions, providing valuable insights into their performance and reliability.