The properties and applications of single-molecule DNA sequencing

John F Thompson,Patrice M Milos

doi:10.1186/gb-2011-12-2-217

John F Thompson, Patrice M Milos

Open Access

https://doi.org/10.1186/gb-2011-12-2-217

Copy DOI

Journal: Genome Biology	Publication Date: Feb 1, 2011
Citations: 197	License type: cc-by

Affiliation: Helicos Biosciences (United States)

Abstract

Single-molecule sequencing enables DNA or RNA to be sequenced directly from biological samples, making it well-suited for diagnostic and clinical applications. Here we review the properties and applications of this rapidly evolving and promising technology.

Highlights

Classical DNA sequencing was developed in the late 1970s and evolved from a low-throughput, almost ‘artisan’ approach, in which the same radiolabeled DNA sample was run on a gel with one lane for each nucleotide [1,2], to an automated method in which all four fluorescently labeled dye terminators for a single sample [3] were loaded onto individual capillaries. ese capillary-based instruments, introduced in 1998, could handle hundreds of individual samples per week, in a manner sufficiently powerful that the first draft sequence of a human genome was finished in 2001 using this technology
The second generation technologies were initially inferior to classical sequencing in terms of read length (about 35 nucleotides for Illumina versus about 700 nt for classical sequencing) and single-read error rate, these shortcomings could be overcome by the sheer volume of data
Continuous improvements in sequencing chemistry have narrowed the gap with respect to read length and errors, as exemplified by Roche 454 routinely achieving read lengths of 400 nt at >99% accuracy [9] and Illumina moving from an initial read length of 36 nt to the current 76 nt or more and raw error rates well below 1%. ese technologies have allowed DNA sequencing to move beyond a method for

Summary

Introduction

Classical DNA sequencing (sometimes referred to as first generation sequencing) was developed in the late 1970s and evolved from a low-throughput, almost ‘artisan’ approach, in which the same radiolabeled DNA sample was run on a gel with one lane for each nucleotide [1,2], to an automated method in which all four fluorescently labeled dye terminators for a single sample [3] were loaded onto individual capillaries. ese capillary-based instruments, introduced in 1998, could handle hundreds of individual samples per week, in a manner sufficiently powerful that the first draft sequence of a human genome was finished in 2001 using this technology. The second generation technologies were initially inferior to classical sequencing in terms of read length (about 35 nucleotides (nt) for Illumina versus about 700 nt for classical sequencing) and single-read error rate (about 2% versus less than 0.1%), these shortcomings could be overcome by the sheer volume of data.

Results

Conclusion