Length Matters: Toward the Clinical Adoption of Long-Read Sequencing

Next-generation sequencing (NGS) has been rapidly integrated into clinical laboratory practice [1]. Its applications continue to expand across diverse areas, including the diagnosis of rare diseases, circulating tumor DNA detection and minimal residual disease monitoring in solid cancers and hematological malignancies, HLA immunogenetics, pharmacogenetics, and clinical microbiology [2–8]. Despite its transformative impact on clinical genomics, conventional short-read NGS, which typically interrogates only a few hundred base pairs of contiguous sequence, has intrinsic limitations. Many clinically important genomic features remain difficult (or even impossible) to resolve accurately using short reads alone. These include structural variants, tandem repeat expansions, and highly repetitive genomic regions.

Long-read sequencing (LRS) has emerged as a solution to these limitations. By generating continuous reads spanning tens of kilobases, LRS can capture the genomic features that are inaccessible or incompletely resolved by short-read sequencing [9]. Over the past decade, it has evolved from an error-prone research technology to a clinically viable platform. Two major platforms have emerged in the field: single-molecule real-time (SMRT) sequencing, developed by Pacific Biosciences (PacBio), and nanopore sequencing from Oxford Nanopore Technologies. Both routinely generate reads spanning tens of kilobases [10]. The advent of circular consensus “HiFi” reads has endowed SMRT sequencing with per-base accuracy close to that of short-read platforms while preserving long read lengths, and enabled the assembly of the first complete telomere-to-telomere human reference genome [11, 12]. Consequently, PacBio HiFi sequencing has emerged as a preferred platform when both accuracy and read length are critical, and its applications in laboratory medicine continue to expand [13]. Oxford Nanopore sequencing offers complementary strengths, including ultra-long reads, direct detection of base modifications without chemical conversion, and rapid, real-time sequencing suited to time-critical applications. Several advances explain why clinical adoption is accelerating: HiFi chemistry has brought per-base accuracy close to that of short reads [10], while maturing analytical pipelines and systematic benchmarking have defined optimal strategies for variant detection [9, 12]. Together, these advances improve workflow robustness and support accumulating clinical validation.

The applications of LRS include structural variant detection, haplotype resolution through phasing, native, conversion-free DNA methylation profiling with allele-specific resolution, and accurate characterization of repetitive genomic regions. Among these applications, tandem repeat analysis is of particular clinical importance. Expansions of short tandem repeats underlie a growing group of neurodegenerative and neuromuscular disorders, for which accurate repeat sizing directly informs diagnosis, prognosis, and disease monitoring [14, 15]. Because these loci are often GC-rich, highly repetitive, and structurally complex, they precisely represent the genomic regions where conventional methods, such as Southern blotting, repeat-primed PCR, and short-read sequencing, perform poorly, making them particularly well suited for LRS. Several target enrichment strategies have been developed to improve the efficiency and scalability of LRS. PCR-based amplicon methods are straightforward but struggle with long, GC-rich repeats, whereas hybridization capture depends on DNA fragmentation, which sacrifices long-range contiguity and enriches expanded, GC-rich loci unevenly. Nanopore adaptive sampling enables computational target enrichment without additional wet-lab procedures; however, its enrichment efficiency and sequencing depth remain limited [16].

In this issue of Annals of Laboratory Medicine, Yeom et al. [17] evaluated the PacBio PureTarget repeat-expansion panel using eight well-characterized reference samples from the Coriell Institute and six patient samples previously tested for FMR1 expansion via repeat-primed PCR, along with replicate normal-control libraries. The panel simultaneously targets 20 clinically relevant repeat-expansion genes. It uses clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein 9-mediated excision for amplification-free target enrichment followed by HiFi sequencing, thereby avoiding amplification bias. Libraries were sequenced on the PacBio Sequel II system, and repeats were genotyped using the tandem repeat genotyping tool. The results were benchmarked against repeat-primed PCR and Southern blotting. In their study, targeted LRS demonstrated complete concordance with repeat-primed PCR and Southern blotting across all samples, while providing clinically relevant information beyond that obtainable by conventional assays. While electrophoretic methods could only determine that FMR1 alleles exceeded 200 repeats, the panel provided precise repeat counts and additionally identified a TCF4 expansion that likely would have been missed by single-locus testing. Its ability to interrogate 20 genes simultaneously is particularly valuable for disorders with overlapping phenotypes, such as spinocerebellar ataxias, where a single assay can replace sequential single-gene testing. The method also enabled DNA methylation profiling within the same workflow. Collectively, these advantages, including accurate sizing of long repeat expansions, multiplex testing, phasing, and epigenetic characterization, address many of the inherent limitations of Southern blotting, repeat-primed PCR, and short-read NGS.

Despite its unique strengths, LRS faces several hurdles that limit its routine adoption in clinical laboratories. LRS remains less cost-effective than established short-read approaches because of its labor-intensive workflows, lower throughput, and requirements for high-quality, high-molecular-weight DNA [18]. Although its accuracy has improved markedly, the complete replacement of short-read sequencing remains premature [12, 18]. Hence, the near-term challenge is to define the complementary roles of long- and short-read sequencing and to establish reimbursement frameworks that support their combined use. Nevertheless, the clinical value of LRS is increasingly evident, as demonstrated in the study by Yeom et al. and in other studies [13, 17, 19]. As the challenges mentioned above are addressed, LRS is likely to evolve from a complementary technology into a front-line diagnostic tool in clinical laboratories.