Leveraging long read sequencing from a single individual to provide a comprehensive resource for benchmarking variant...

Area of Science:

• Genomics
• Bioinformatics

Background:

• High-throughput sequencing necessitates accurate human variant sets for precision medicine.
• Existing variant sets lack comprehensive inclusion of all variant types, particularly structural variants (SVs).

Purpose of the Study:

• To construct a high-confidence, comprehensive human variant gold set from the HuRef genome.
• To address limitations in previous variant datasets for benchmarking sequencing algorithms.

Main Methods:

• Leveraged high-quality Sanger sequences from the HuRef genome.
• Cross-validated variants with deep Illumina sequencing, population datasets, and established algorithms.
• Reanalyzed the HuRef genome to resolve compatibility, organization, and accuracy issues.

Main Results:

• Developed a comprehensive gold set of human variants, including small variants and large deletion/insertion SVs (up to 100kb).
• Achieved high specificity and sensitivity in the constructed gold set.
• Demonstrated the utility of the HuRef gold set for benchmarking SV detection tools.

Conclusions:

• The HuRef gold set represents a significant advancement in human variant resources.
• This resource enables more accurate assessment of sequencing algorithms for precision medicine applications.
• The comprehensive nature of this gold set, including large SVs, fills a critical gap in current genomic benchmarking tools.