Optimizing GRIDSS for clinical use: A targeted NGS filtering strategy for germline structural variant detection.

Area of Science:

• Genomics
• Molecular Diagnostics
• Bioinformatics

Background:

• Detecting intermediate-sized structural variants (SVs) is crucial for diagnosing hereditary conditions but remains a challenge for current diagnostic tools.
• Existing methods often fall short, necessitating advanced approaches for comprehensive variant detection.

Purpose of the Study:

• To evaluate the utility of the GRIDSS (Genomic Rearrangement Identification and Detection by Sequence Synthesis) tool for detecting germline structural variants (SVs) in a large cohort of hereditary cancer patients.
• To develop and optimize a filtering strategy for prioritizing clinically relevant SVs within a diagnostic workflow.

Main Methods:

• Analyzed next-generation sequencing (NGS) panel data from 9726 patients with suspected hereditary cancer using GRIDSS.
• Integrated paired-end mapping, split-read analysis, and assembly-based approaches within the GRIDSS framework.
• Developed and applied a filtering strategy, including parameter optimization and visual inspection, followed by Sanger/Nanopore long-read sequencing for validation.

Main Results:

• Successfully reduced over a million variants to 89 candidates using the developed filtering strategy.
• Confirmed 13 likely true positive germline SVs, with all experimentally validated.
• Identified 8 (likely) pathogenic variants, including frameshift duplications, splicing variants, and mobile element insertions in key cancer predisposition genes (MSH6, BARD1, APC, BRCA2, PALB2).

Conclusions:

• Implementation of GRIDSS in a diagnostic setting significantly increased the diagnostic yield for detecting clinically relevant germline SVs.
• The developed workflow demonstrates feasibility for routine diagnostic use, offering a comprehensive approach to germline SV detection.