Accurate whole human genome sequencing using reversible terminator chemistry

Abstract

DNA sequence information underpins genetic research, enabling discoveries of important biological or medical benefit. Sequencing projects have traditionally employed long (400–800 bp) reads, but the existence of reference sequences for the human and many other genomes makes it possible to develop new, fast approaches to re-sequencing, whereby shorter reads are compared to a reference to identify intra-species genetic variation. We report an approach that generates several billion bases of accurate nucleotide sequence per experiment at low cost. Single molecules of DNA are attached to a flat surface, amplified in situ and used as templates for synthetic sequencing with fluorescent reversible terminator deoxyribonucleotides. Images of the surface are analysed to generate high quality sequence. We demonstrate application of this approach to human genome sequencing on flow-sorted X chromosomes and then scale the approach to determine the genome sequence of a male Yoruba from Ibadan, Nigeria. We build an accurate consensus sequence from >30x average depth of paired 35-base reads. We characterise four million SNPs and four hundred thousand structural variants, many of which are previously unknown. Our approach is effective for accurate, rapid and economical whole genome re-sequencing and many other biomedical applications.

Direct mutation analysis by high-throughput sequencing: From germline to low-abundant, somatic variants

“…There are currently three routinely used platforms for MPS: the Roche 454 system[58], the Illumina HiSeq[59], and the SOLiD system of Applied Biosystems[60]. A fourth, the PacBio RS single molecule sequencing system of Pacific Biosciences[61], was introduced fairly recently and will be briefly discussed later.…”

“…The two major alignment algorithms used are hash table–based algorithms (BLAST, MAQ[68], Eland[59]) and Burrows Wheeler transform (BWT)-based methods (SOAP2[69], BWA[70], Bowtie[71]). Hash table-based algorithms use an indexing scheme that enables ultra-fast searches for short sequences of a defined length k (k-mer matches or seeds) with up to m mismatches.…”

Optimized filtering reduces the error rate in detecting genomic variants by short-read sequencing

“…Another commonly used approach is to apply quality filters that are aimed at selectively removing errors. Every whole-genome sequence reported so far has used filtering to some extent: the most commonly used filters being those that remove sequences with a too-low coverage depth, discard variants with a low-confidence score or eliminate variants located within a cluster of variants 3,7,[10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25] . Surprisingly, there is little consensus with respect to which filters should be used and at which threshold they should be applied.…”

The pursuit of genome-wide association studies: where are we now?

“…So far, seven human diploid genomes have been fully sequenced, and some important insights have been gained from these resequencing studies. [145][146][147][148][149][150][151] The most prominent finding from these studies is that, besides SNPs, other genetic variants are also abundant in the human genome. These studies found that in addition to the 3-4 million SNPs, several hundred-thousand of short indels (for example, sizes defined as 3 and 16 bp or less in the Bentley et al 147 and Wang et al 148 study, respectively) are also present in each individual human genome.…”