By Kevin Davies
September 8, 2009 | Complete Genomics has named Pfizer, Duke University, Brigham & Women’s Hospital, the Ontario Institute for Cancer Research and the Flanders Institute for Biotechnology (VIB) among its first dozen clients for its human genome sequencing service. The company has delivered 14 genomes since March, in projects ranging from cancer to HIV susceptibility to schizophrenia.
Those organizations join the Broad Institute, the Institute of Systems Biology, and the Hudson Alpha Institute as early pilot project partners for Complete. The Mountain View, California, sequencing firm is striving to convince sequencing labs, many of which have a massive investment of resources and personnel in genome sequencing, that delivering completed genome sequences can offer at least a partial alternative to doing the work themselves.
Notable in the group of partners announced today is Pfizer, the first big pharma company that Complete has announced. CEO Clifford Reid stated last year that big pharma was a key market for the company, however Pfizer has not offered any further details on its participation.
Complete Genomics has also sequenced the genome of one member of the Personal Genome Project (PGP), founded by George Church, a Complete Genomics scientific advisor and professor of genetics at Harvard Medical School. Church told Bio-IT World that his group is in the midst of sequencing the whole genomes for the first ten PGP volunteers, the PGP-10, all of whom have already undergone full exon sequencing.
“It looks like their quality is really quite high,” said Church of the PGP genome sequenced by Complete Genomics. “The claim of 10-5 [error rate] seems reasonable, which is not completely out of line with other high-quality projects.” Church said they were aiming to do high coverage, in order to identify both maternal and paternal alleles. “If you do 7x coverage, which was the gold standard for many years throughout most of the genome project and the first few diploid genomes, that really doesn’t give you adequate coverage.”
Church said his group together with Complete Genomics has submitted a paper that describes the sequencing of 14 human genomes and the detailed analysis of three of them. “We’ve looked at putative trait-altering alleles, using software called Trait-o-matic, mentioned briefly in the Nature paper on the Korean genome,” said Church. “It really enriches for putative mutations that are likely to be sequencing errors. In other words, if you find what looks like an incredibly deleterious mutation by various criteria, and the person is perfectly healthy and walking around, the chances are it’s a sequencing error. So we use that to prioritize and try to find sequencing errors, in addition to other quality assessment tools.”
Church added that he was very excited by the potential of Complete Genomics’ long-fragment read approach, which he said “is the best way right now – not forever, but right now – to get haplotyping and long-distance sequence information.”
As previously announced, Complete is charging as little as $5,000 per genome for high-volume customers, but $20,000 per genome for smaller orders (minimum eight genomes). Complete plans its full commercial launch in January 2010, beginning an ambitious program to sequence 10,000 human genomes next year using new commercial-scale instruments capable of reading multiple arrays simultaneously, and producing more than one terabase per run.
Complete is also developing higher-density arrays containing 2.85 billion spots of its proprietary DNA nanoballs, laid out in a grid. The company aims to sequence an entire human genome on a single array. At a throughput of one instrument-day per genome, a fleet of fewer than 30 instruments should in theory generate the stated annual goal of 10,000 genomes (assuming single arrays).
Although the Church lab is known for developing new sequencing technologies, including the basis for the Polonator, Church is optimistic about the long-term prospects for a genome sequencing service model. “To a first approximation -- barring the portable instant sequencer or the sequencer that does various other projects -- if the service is cheaper than what you can do running it in house, you’re going to do the service, because it’s such a hassle to set aside space and personnel. It completely dilutes out everything else you’re trying to do,” he said.