A national long-read sequencing study on chromosomal rearrangements uncovers hidden complexities

Abstract

Clinical genetic laboratories often require a comprehensive analysis of chromosomal rearrangements/structural variants (SVs), from large events like translocations and inversions to supernumerary ring/marker chromosomes and small deletions or duplications. Understanding the complexity of these events and their clinical consequences requires pinpointing breakpoint junctions and resolving the derivative chromosome structure. This task often surpasses the capabilities of short-read sequencing technologies. In contrast, long-read sequencing techniques present a compelling alternative for clinical diagnostics. Here, Genomic Medicine Sweden-Rare Diseases has explored the utility of HiFi Revio long-read genome sequencing (lrGS) for digital karyotyping of SVs nationwide. The 16 samples from 13 families were collected from all Swedish healthcare regions. Prior investigations had identified 16 SVs, ranging from simple to complex rearrangements, including inversions, translocations, and copy number variants. We have established a national pipeline and a shared variant database for variant calling and filtering. Using lrGS, 14 of the 16 known SVs are detected. Of these, 13 are mapped at nucleotide resolution, and one complex rearrangement is only visible by read depth. Two Chromosome 21 rearrangements, one mosaic, remain undetected. Average read lengths are 8.3-18.8 kb with coverage exceeding 20× for all samples. De novo assembly results in a limited number of phased contigs per individual (N50 6-86 Mb), enabling direct characterization of the chromosomal rearrangements. In a national pilot study, we demonstrate the utility of HiFi Revio lrGS for analyzing chromosomal rearrangements. Based on our results, we propose a 5-year plan to expand lrGS use for rare disease diagnostics in Sweden.

Figure 1.

Quality measures of lrGS. (A) Read length distribution, (B) coverage, and (C) N50 of de novo assembly.

Figure 2.

Translocations and inversions identified with lrGS. Circos plots of rearrangements detected in four cases using lrGS: a t(4;9) in P1, a t(1;10) and inv(2) (P7.1 and P7.3), a t(X;9) (P9), and a t(1;4;6;4) (P12). A green/red line indicates copy number gain/loss, respectively. Genes disrupted are indicated at the breakpoint.

Figure 3.

Subway plots of two complex rearrangements: (A) a complex dup-trip-quad-trip-dup-del on Chromosome 2 observed in P4 and (B) a clustered CNV on Chromosome 3 detected in P5. Deleted segments are shown in red, and arrows mark inverted segments.

Figure 4.

Characterization of background SVs. (A) Boxplot illustrating the number of SVs per type (common in yellow and rare in green). (B) Violin plot of SV length per SV type (excluding SV > 5 kbp). (C) Allele frequency histogram. (D) Comparison of allele frequencies between the GMS long-read cohort and SweGen srGS SV database (common in yellow, missing in gray, and rare in green).

Figure 5.

Toward lrGS as a first tier diagnostic test in rare disease. A 5-year time line with the expected development of long-read sequencing in the clinical setting.