講演情報
[OE2-4]Scalable and accurate detection of technically challenging variants through modified next-generation sequencing
○Aradhya Swaroop, Rojahn Susan, Truty Rebecca, Kennemer Michael, Paul Joshua (Invitae Inc., San Francisco, USA)
Introduction: NGS predominates genetic testing but some genes and variants are not readily evaluated due to technical challenges created by pseudogenes (i.e., PMS2, SMN1/2, GBA, HBA1/HBA2, CYP21A2), repetitive sequences (ARX-poly Alanine, FMR1-AGG, CFTR poly T/TG), and other complexities (MSH2-Boland inversion). Traditional methods can address these challenges, but are often laborious and expensive, and unsuitable for large-scale testing. NGS workflows can be customized to overcome these challenges.
Materials and Methods: We customized our NGS assay and bioinformatics processes to detect each class of technically challenging variant. Adaptations included target-sequence enrichment, split-read detection, and bioinformatic masking of highly similar sequences. The customizations were validated by orthogonal methods.
Results: Each technically challenging gene or variant required bespoke modifications to the NGS workflow. For some, NGS provided a high-sensitivity screen followed with a confirmatory assay. For others, the NGS modifications themselves were sufficient. Validation concordance with reference samples was 100% for PMS2, SMN1/2, GBA, HBA1/2, FMR1-AGG, and CFTR poly T/TG; 86.3% for ARX-polyA; 97.9% for CYP21A2; and 99.9% MSH2-Boland.
Conclusions: Technically challenging genes and variants can be evaluated with custom NGS tests for greater efficiency. Automated lab processes, high-depth sequence coverage, and tailored bioinformatics are critical. This ability supports a path to identify all clinically important variant types from a single assay with a single sample.
Materials and Methods: We customized our NGS assay and bioinformatics processes to detect each class of technically challenging variant. Adaptations included target-sequence enrichment, split-read detection, and bioinformatic masking of highly similar sequences. The customizations were validated by orthogonal methods.
Results: Each technically challenging gene or variant required bespoke modifications to the NGS workflow. For some, NGS provided a high-sensitivity screen followed with a confirmatory assay. For others, the NGS modifications themselves were sufficient. Validation concordance with reference samples was 100% for PMS2, SMN1/2, GBA, HBA1/2, FMR1-AGG, and CFTR poly T/TG; 86.3% for ARX-polyA; 97.9% for CYP21A2; and 99.9% MSH2-Boland.
Conclusions: Technically challenging genes and variants can be evaluated with custom NGS tests for greater efficiency. Automated lab processes, high-depth sequence coverage, and tailored bioinformatics are critical. This ability supports a path to identify all clinically important variant types from a single assay with a single sample.