DNA sequencing is currently dominated by short-read platforms from Illumina (and to a lesser extent Ion Torrent, from Waltham, Massachusetts–based ThermoFisher, which have enabled faster and cheaper sequencing. But some scientists are bumping up against the limits of standard short-read technology, particularly for de novo assembly (mapping whole genomes from scratch), haplotyping (distinguishing between maternal and paternal copies of each chromosome) or defining structural variation.
Short-read sequencers generate multitudinous reads that each span only a few hundred bases, which must then be computationally reassembled into larger 'scaffolds' of genomic data based on the overlap between individual reads. But this software-driven assembly is often confounded by long stretches of repetitive sequences or large structural rearrangements, making it difficult or impossible to assign locations for some reads and introducing errors. “99% of the complex insertions and short tandem repeats and variable-number tandem repeats are missed with the Illumina technology,” says Jonas Korlach, CSO of Menlo Park, California–based Pacific Biosciences (PacBio), referring to a recent study by Evan Eichler of the University of Washington in Seattle, which made use of the company's long-read technology to map structural rearrangements in cancer (Nature 517, 608–611, 2015).
This is a preview of subscription content, access via your institution