Expectations and blind spots for structural variation detection from long-read assemblies and short-read genome...

Area of Science:

• Genomics
• Bioinformatics
• Human Genetics

Background:

• Short-read whole-genome sequencing (srWGS) is the standard for large-scale genomic studies but struggles with structural variant (SV) detection.
• Long-read whole-genome sequencing (lrWGS) technologies are emerging as a powerful alternative for comprehensive genomic analysis.

Purpose of the Study:

• To compare SV detection capabilities between srWGS and lrWGS.
• To quantify the genomic properties and unique contributions of SVs identified by each technology.
• To establish expectations for routine SV detection using srWGS.

Main Methods:

• Comparative analysis of SVs detected by srWGS and lrWGS assembly in three families from the Human Genome Structural Variation Consortium (HGSVC).
• Evaluation of SV detection power and precision across different genomic contexts and variant classes.
• Assessment of SVs in segmental duplication (SD) and simple repeat (SR) regions versus non-SD/SR regions.

Main Results:

• lrWGS assembly identified approximately twice as many SVs per genome (~25,000) compared to srWGS (~11,000).
• 91.4% of deletions uniquely identified by lrWGS were located in segmental duplication (SD) and simple repeat (SR) regions, which constitute 9.7% of the reference genome.
• High concordance (93.8%) for deletions was observed between srWGS and lrWGS in non-SD/SR regions, while lrWGS showed superior detection of insertions across all contexts.

Conclusions:

• While srWGS provides high concordance for deletions in non-repetitive regions, lrWGS significantly enhances the detection of insertions and SVs within complex genomic areas like SDs and SRs.
• Assembly-based lrWGS is crucial for creating comprehensive catalogs of insertions, transposable elements, and repeat expansions, particularly in disease-associated regions previously difficult to assess.
• The findings suggest incremental improvements for pathogenic deletion detection in disease-associated exons using lrWGS, but highlight its substantial value for novel variant discovery in recalcitrant genomic sequences.