The $1,000 Genome: Game Changer or PR Stunt?
Posted by Jessica Cussins February 6, 2014
Biopolitical Times
On January 14, the preeminent DNA sequencing company Illumina revealed its HiSeq X Ten, a machine that can reportedly sequence an entire human genome in a few hours, for under $1,000. The announcement of a “$1,000 genome” has long been anticipated as an important rite of passage for the field. In 2008, the same feat cost $350,000; the $1,000 benchmark is supposed to represent the point at which whole genome sequencing will move into the mainstream and trigger a “revolution in personalized medicine.” For now, however, its significance may be found primarily in its creation of good PR for Illumina.
The $1,000 genome era has been coaxed forward for over a decade. The federal government alone has poured tens of millions of dollars into grants. In 2002, Craig Venter’s foundation offered $500,000 to the first team that could reach the $1,000 benchmark; the Archon X Prize competition later raised that offer to $10 million. Illumina won’t actually receive any compensation because the competition was canceled in August, but the company is getting its fair share of recognition: “They are brilliant… Well ahead of the field,” “It’s a terrific company, and they are in a very dominant position,” etc.
However, the $1,000 price tag is actually somewhat misleading. The machine itself costs $1 million, and Illumina will only sell you a minimum of 10 machines. That’s $10 million upfront, so you have to be sequencing a lot of people to bring the cost down. As computational biologist Mick Watson put it, "I think they might be right in claiming the $1000 genome - if you do 18,000 human genomes per year for four years on each X Ten system. That's a lot of human genomes though." Furthermore, it turns out that this cost probably doesn’t include realistic labor costs, analysis, or commercial markup for providers of the test (see ‘Update’.) And finally, the price obviously does not include potentially substantial downstream costs such as over-testing and false positives.
Many people have argued that the real challenge with genetic sequencing is not in the sequencing itself, but in the interpretation and understanding of the results. A lot of the skepticism and criticism of the personal genomics field in the past few months has centered on exactly this point.
The PHG Foundation, an independent non-profit UK organization whose tag line is “making science work for health” recently released a new briefing note on whole genome analysis that highlighted a number of concerns with clinical applications. Most notably, PHG questions whether the lack of standardization in the interpretation of results rules out the possibility of information that is actually reliable or useful. The note mentions that one recent study found that 27% of the disease-associated mutations described in the published literature are incorrect, and another found a “43% mismatch between variants identified from a single sample processed by five common alignment and variant calling pipelines.” The note concludes that challenges to whole genome sequencing are not insurmountable, but that they will require care, time and resources to overcome.
Naming similar concerns, but coming to a somewhat different conclusion, an article in The Scientist last year argued that whole genome sequencing should not be used indiscriminately for all patients, even if the price does go down. The authors wrote, “in spite of an understandable narcissism regarding our own genomes, the reality for most of us is that our genomes are incredibly boring. Given our current understanding of how to interpret—much less apply—genomic information, the average person’s genome yields precious little knowledge that will lead to better health.”
Additionally, they continue,
All this considered, it is clear that the arrival of the $1,000 genome will not, on its own, imminently herald an age of personalized medicine. And although catchphrase-worthy goals can encourage speedy innovation, they can also overshadow pesky points such as the actual usefulness or desirability of the products themselves.
Previously on Biopolitical Times:
The $1,000 genome era has been coaxed forward for over a decade. The federal government alone has poured tens of millions of dollars into grants. In 2002, Craig Venter’s foundation offered $500,000 to the first team that could reach the $1,000 benchmark; the Archon X Prize competition later raised that offer to $10 million. Illumina won’t actually receive any compensation because the competition was canceled in August, but the company is getting its fair share of recognition: “They are brilliant… Well ahead of the field,” “It’s a terrific company, and they are in a very dominant position,” etc.
However, the $1,000 price tag is actually somewhat misleading. The machine itself costs $1 million, and Illumina will only sell you a minimum of 10 machines. That’s $10 million upfront, so you have to be sequencing a lot of people to bring the cost down. As computational biologist Mick Watson put it, "I think they might be right in claiming the $1000 genome - if you do 18,000 human genomes per year for four years on each X Ten system. That's a lot of human genomes though." Furthermore, it turns out that this cost probably doesn’t include realistic labor costs, analysis, or commercial markup for providers of the test (see ‘Update’.) And finally, the price obviously does not include potentially substantial downstream costs such as over-testing and false positives.
Many people have argued that the real challenge with genetic sequencing is not in the sequencing itself, but in the interpretation and understanding of the results. A lot of the skepticism and criticism of the personal genomics field in the past few months has centered on exactly this point.
The PHG Foundation, an independent non-profit UK organization whose tag line is “making science work for health” recently released a new briefing note on whole genome analysis that highlighted a number of concerns with clinical applications. Most notably, PHG questions whether the lack of standardization in the interpretation of results rules out the possibility of information that is actually reliable or useful. The note mentions that one recent study found that 27% of the disease-associated mutations described in the published literature are incorrect, and another found a “43% mismatch between variants identified from a single sample processed by five common alignment and variant calling pipelines.” The note concludes that challenges to whole genome sequencing are not insurmountable, but that they will require care, time and resources to overcome.
Naming similar concerns, but coming to a somewhat different conclusion, an article in The Scientist last year argued that whole genome sequencing should not be used indiscriminately for all patients, even if the price does go down. The authors wrote, “in spite of an understandable narcissism regarding our own genomes, the reality for most of us is that our genomes are incredibly boring. Given our current understanding of how to interpret—much less apply—genomic information, the average person’s genome yields precious little knowledge that will lead to better health.”
Additionally, they continue,
~1 percent of us may find our genomes to hold rather terrifying information—mutations that strongly predispose us to disturbing diseases for which we currently have no preventive modalities and no treatments. Contrary to the usual trope that most people will “want to know everything” in their genome, actual data indicate that most people who are at risk of having mutations in such genes do not wish to know. We need to think very carefully before assuming that widespread genome sequencing will be universally useful or even desirable.But the companies and universities that buy whole-genome sequencing machines will have a strong incentive to use them for many people, and to translate that mountain of data into information that at least seems meaningful and useful. One danger of so much front-end investment is that it could skew research interests toward a search for genetic explanations for diseases and traits that in fact have other etiologies. The causes and trajectories of most common diseases and traits are so integrated with environmental, social and lifestyle factors that an over-emphasis on genes could actually be counterproductive. Additionally, increased use of whole genome sequencing could encourage researchers, and marketing efforts, to promote theories of genetic determinism, potentially encouraging the disturbing resurgence of biological explanations for social ills.
All this considered, it is clear that the arrival of the $1,000 genome will not, on its own, imminently herald an age of personalized medicine. And although catchphrase-worthy goals can encourage speedy innovation, they can also overshadow pesky points such as the actual usefulness or desirability of the products themselves.
Previously on Biopolitical Times: