Closing the gaps, and improving somatic structural variant analysis and benchmarking using CHM13-T2T

Area of Science:

• Genomics
• Cancer Research
• Bioinformatics

Background:

• Advancements in sequencing and bioinformatics aid cancer genome interpretation.
• Structural variants (SVs) are key somatic events in tumors.
• Detecting and annotating somatic SVs remains challenging despite long-read sequencing.

Purpose of the Study:

• To evaluate if the completed human reference genome (CHM13-T2T) improves somatic structural variant (SV) calling.
• To develop methods for annotating SVs using CHM13-T2T.
• To establish an updated benchmark for somatic SV calling.

Main Methods:

• Comparative analysis of SV detection using GRCh38 versus CHM13-T2T reference genomes.
• Read alignment to CHM13-T2T followed by coordinate lifting to GRCh38 for annotation.
• Assessment of a structural variant benchmark set across multiple replicates and sequencing technologies.

Main Results:

• CHM13-T2T demonstrated improved SV detection accuracy and reduced false positives compared to GRCh38.
• A method was developed to combine CHM13-T2T alignment with GRCh38 annotations.
• The COLO829/COLO829BL cell line exhibited instability, with a consensus set of 54 stable somatic SVs identified.

Conclusions:

• The CHM13-T2T reference genome enhances somatic SV detection in cancer.
• An updated benchmark set for somatic SV calling, with coordinates for both GRCh38 and CHM13-T2T, has been proposed.
• These approaches offer potential improvements for genetic disease research and cancer diagnostics.