longreadvariants

Longreadvariants refer to genetic variants identified or characterized using long-read sequencing data. These variants include single-nucleotide variants and small insertions or deletions, as well as structural variants such as insertions, deletions, inversions, duplications, and complex rearrangements that are often difficult to resolve with short-read sequencing. The term emphasizes the use of long reads to obtain haplotype-resolved or phased variant information.

Long-read sequencing technologies such as PacBio SMRT and Oxford Nanopore produce reads that span thousands to tens of thousands of bases, enabling variant discovery in repetitive regions and complex loci, and allowing direct observation of large insertions and complex rearrangements. This improves detection of mobile element insertions, tandem duplications, and multi-allelic variants, and supports haplotype phasing across long genomic distances.

Analytical approaches involve aligning reads to a reference (often with minimap2), calling SNVs and small indels with long-read–aware tools (Longshot, Clair), and calling structural variants with tools such as Sniffles, cuteSV, or PBSV. Phasing and haplotype reconstruction can be performed with WhatsHap or HapCUT2, sometimes enabling haplotype-resolved assemblies.

Applications include human genetics research, cancer genomics, and clinical genomics, where longreadvariants can reveal variants hidden by short reads, resolve complex rearrangements, and improve reference-quality genome assemblies. Limitations include historically higher per-base error rates (though improved with HiFi reads), higher cost, larger data volumes, and greater computational demands.

See also: long-read sequencing, structural variant, haplotype phasing.