HiFi long-read genomes for difficult-to-detect, clinically relevant variants.

Area of Science:

• Genomics
• Medical Genetics

Background:

• Short-read sequencing is standard for rare disease diagnostics but struggles with complex genomic regions.
• Existing methods often require multiple complementary tests to detect all relevant variants.

Purpose of the Study:

• To evaluate long-read sequencing (LRS) as a potential first-tier diagnostic test for rare diseases.
• To assess LRS accuracy in detecting challenging germline variants missed by short-read sequencing.

Main Methods:

• Performed 30× high-fidelity (HiFi) long-read sequencing on 100 samples with 145 known difficult-to-detect variants.
• Utilized standard variant callers to analyze sequencing data.
• Conducted titration analysis to assess performance at reduced coverage.

Main Results:

• LRS successfully re-identified 83% of variants automatically, including structural variants, indels in homologous regions, and short tandem repeat expansions.
• An additional 10% of variants were visually apparent but not automatically called.
• Systematic challenges were noted for specific variant types, like AG-rich repeat expansions.
• 93% of challenging pathogenic variants were identified, with 90% of variants detectable at 15× coverage.

Conclusions:

• Long-read sequencing demonstrates high accuracy in detecting a wide spectrum of clinically relevant rare disease variants.
• LRS shows potential as a unified, cost-effective diagnostic approach, reducing the need for multiple testing modalities.
• Further optimization is needed for specific complex variant types, but LRS is a significant advancement for rare disease diagnostics.