Mar 08 2012
MOUNTAIN VIEW, California — In Silicon Valley, the line between computing and biology has begun to blur in a way that could have enormous consequences for human longevity.
Bill Banyai, an optical physicist at Complete Genomics, has helped make that happen. When he began developing a gene sequencing machine, he relied heavily on his background at two computer networking start-up companies. His digital expertise was essential in designing a factory that automated and greatly lowered the cost of mapping the three billion base pairs that form the human genome.
The promise is that low-cost gene sequencing will lead to a new era of personalized medicine, yielding new approaches for treating cancers and other serious diseases. The arrival of such cures has been glacial, however, although the human genome was originally sequenced more than a decade ago.
Now that is changing, in large part because of the same semiconductor industry manufacturing trends that opened up consumer devices like the PC and the smartphone: exponential increases in processing power and transistor density are accompanied by costs that fall at an accelerating rate.
As a result, both new understanding and new medicines will arrive at a quickening pace, according to the biologists and computer scientists.
“For all of human history, humans have not had the readout of the software that makes them alive,” said Larry Smarr, director of the California Institute of Telecommunications and Information Technology, a research center that is jointly operated by the University of California, San Diego, and the University of California, Irvine, who is a member of the Complete Genomics scientific advisory board. “Once you make the transition from a data poor to data rich environment, everything changes.”
Complete Genomics, based in Mountain View, is one of more than three dozen firms hastening to push the cost of sequencing an entire human genome below $1,000. The challenge is part biology, part chemistry, part computing, and in Complete Genomics’ case, part computer networking.
Complete Genomics is a classic Silicon Valley start-up story. Even the gene sequencing machines, which are housed in a 4,000-square-foot room bathed in an eerie blue light, appear more like a traditional data center than a biology lab.
In 2005 ,when Clifford Reid, a successful Silicon Valley software entrepreneur, began to assemble his team, he approached Dr. Banyai and asked if he was interested in joining a gene sequencing start-up. Dr. Reid, who was also trained in physics and math, had spent a year as an entrepreneur-in-residence at the Massachusetts Institute of Technology, where he had become a convert to bioinformatics, the application of computer science and information technologies to biology and medicine.
Dr. Banyai had even less experience in biology. Formerly with the Internet networking start-ups GlimmerGlass and Silicon Light Machines, he in turn began by reading a pioneering 2005 article in the journal Science in which a group of researchers in George Church’s genetics laboratory at Harvard describe a new technique intended to speed gene sequencing.
Today Dr. Banyai is finishing the second generation of a machine that blends robotics, chemistry, optics and computing. It is emblematic of the serendipitous changes that take place when a manufacturing process is transformed: performance increases and cost falls at an accelerating rate.
“Genomes are now being sequenced incredibly cheaply,” said Russ B. Altman, who is a founder of Personalis, a start-up based in Palo Alto, Calif., that is developing software to interpret genomes. “On the discovery and science side we will be able to do clinical trials. We’ll be able to check the entire genome.”
Recently, on the company’s Web site, Dr. Reid predicted that the cost of gene sequencing could eventually be as low as that of a blood test: “I believe that the impact on the medical community of whole human genome sequencing at a cost comparable to a comprehensive blood test will be profound, and it will raise a host of public policy issues (privacy, security, disclosure, reimbursement, interpretation, counseling, etc.), all important topics for future discussions,” he wrote.
Dr. Banyai said he had found that Silicon Valley start-up ideas tracked well. “There is this remarkable thing that happens in start-ups. You make up this plan and then you step off a cliff and magically a little bridge appears,” he noted, as new technologies appear in the nick of time.
In the case of Complete Genomics, the company is riding in part on big advances being made in industrial digital cameras that are capable of capturing the fluorescent molecules that are used to “read” small sequences of DNA.
In the last half-year, a new generation of cameras, more frequently used for factory inspection systems, has made it possible to speed up the Complete Genomics sequencing process tenfold. That, the company has said, will drive its capacity to 100,000 genomes annually from 10,000.
The parallels between the evolution of the nascent gene sequencing industry and the Valley’s chip makers are striking. By placing more circuits on a silicon wafer at an exponentially increasing pace since the early 1960s, the semiconductor industry transformed the cost of computing. As a result, today the world’s most powerful supercomputer from the 1980s nestles comfortably in your hand and costs several hundred dollars.
Complete Genomics’ competitors are also exploiting designs to drive costs down. For example, Life Technologies, based in Carlsbad, Calif. uses a direct approach to read the bases in the genome from an array of sensors on the surface of a semiconductor chip. As more sensors are packed onto each successive generation of technology, the cost of sequencing will also fall sharply.
Last month, Oxford Nanopore Technologies created an industry sensation when it introduced a machine that sequenced genes using an alternative approach called nanopore sequencing, in which a strand of DNA is read as it is pulled through a microscopic hole.
The system is scheduled to be available later this year. However, it has an error rate much higher than that of the Complete Genomics system, which has independently been given high marks for accuracy.
Because there is no clear winner yet, all of the companies are pushing hard to get down the cost curve as fast as possible
In 2011, Complete Genomics became one of the market leaders. This year, it has produced more than 3,000 sequences at a cost of about $5,000 each. Dr. Banyai’s higher capacity second generation system is now being installed and will begin production during the first half of this year. A third generation design has been completed.
What initially set Complete Genomics apart from the field was its strategy of offering gene sequencing as a service, rather than selling a machine to laboratories. More recently, Illumina, one of its crucial competitors, has also begun offering sequencing as a service, in addition to selling its machines.
“Our competitors have to supply kits that can be executed by a graduate student rolling out of bed with a hangover,” said Dr. Reid. “We don’t live with that standard, and that can be tremendously liberating. Ours can be horrifically complex as long as it can be executed by a robot.”
The company also began with the business intent of sequencing only the human genome, rather than those of other species, too — a strategy that was heresy in 2005, when the founders set out to raise money. At that time, only two human genomes had been sequenced. However, Complete Genomics founders argue that focusing just on the human genome has given them a leg up.
“You make a whole bunch of decisions that don’t work well for corn or bacteria, but they work very well for humans,” Dr. Reid said.
John Markoff for The New York Times