One of Complete Genomics' cost advantages comes from its business model. While most sequencing companies sell machines to labs, which then do their own sequencing, Complete Genomics offers a sequencing service. Because everything is done in-house, the company hasn't had to spend time or money making the machines consumer-friendly. And because it focuses only on human genomes, its scientists can optimize the pipeline around this single target.

Lower costs also come from the company's novel method of packing DNA on a chip. Like other, newer sequencing technologies, such as those sold by Illumina and Applied Biosystems, Complete Genomics' technology is based on massively parallel sequencing--it collects thousands of snippets of DNA on a microarray and sequences them at the same time. Strips of DNA are condensed into tight balls, which are then poured onto specially fabricated arrays with unprecedented density--about a billion balls fit on a chip the size of a microscope slide.

"They very efficiently use the real estate on the chip, which saves on camera time and reagent use," says Church, who is on the company's scientific advisory board. Complete Genomics' sequencing chemistry is based on a method developed in Church's lab. (For more on the technology, seeFive Thousand Bucks for Your Genome).

Despite the low reagent costs reported in the paper, Complete Genomics isn't offering $5,000 genomes to the public just yet. "We are still doing genomes in small quantities, like eight, at $20,000 [each]," says Clifford Reid, the company's chief executive officer. "Not all of the cost is materials; some of it is in labor and machines." He says that while larger projects will cost less, "it will take a very large number to get down to $5,000."

Reid anticipates reaching that point in 2010 and 2011. And $20,000 is still cheaper than other offerings--earlier this year, Illumina announced a $48,000 genome sequencing service that could be ordered with a doctor's prescription.

This week Complete Genomics also announced a project with the Institute for Systems Biology, in Seattle, to sequence the genomes of 100 people with Huntington's disease, a rare and debilitating inherited neurological condition, along with their family members. The aim of the project is to find genetic variations that affect the severity of the disease. The study will be the first to search for disease-linked variants across the entire genome--previous studies have used gene microarrays, which survey about a million different spots across the genome.