Diagnostic yield of long-read sequencing for rare diseases: a systematic review

Background: Nearly half of patients with rare genetic disorders remain undiagnosed, which may in part be due to limitations of current short-read sequencing (SRS) approaches in detecting complex genomic alterations. Long-read whole genome sequencing (lrWGS) technologies can address these limitations through enhanced detection of structural variants (SVs), repetitive regions, and epigenetic changes. Methods: To evaluate the diagnostic yield of lrWGS in patients with rare genetic diseases receiving inconclusive or negative results from standard testing, we searched the PubMed, Science Direct, Scopus, and ProQuest databases to July 2025 for studies applying lrWGS to unresolved rare disease cases and reporting diagnostic outcomes. Risk of bias was assessed using the QUADAS-2 tool. Results: Nine studies involving 646 previously unresolved cases that underwent lrWGS met the inclusion criteria. Of these, 29 individuals (24 unique diagnoses involving 25 genes) received a definitive diagnosis through lrWGS, a diagnostic yield of 4.5%. SVs accounted for the majority of identified variants (41.67%), followed by combined SV/single-nucleotide variants (20.83%), methylation changes (16.67%), and other variant types (copy number variations, indels, and tandem repeats). Most detected variants were in regions typically inaccessible to short-read whole-exome sequencing (WES). lrWGS also enabled phasing and methylation analysis in a single assay, which was valuable for compound-heterozygosity detection and diagnostic interpretation. Conclusion: lrWGS shows clear potential for improving diagnostic rates in previously unresolved rare disease cases, particularly when applied after WES and combined with advanced tools such as phasing and methylation profiling. As technologies evolve and become more accessible, lrWGS may increasingly become a first-tier diagnostic approach, especially in phenotypically complex conditions.