Illumina’s CEO on the Promise of the $1,000 Genome — And the Work That Remains
Illumina seized the science world’s attention at the outset of the year by announcing it had achieved the $1,000 genome, crossing a long-sought threshold expected to accelerate advances in research and personalized medicine.
The San Diego company unveiled the HiSeqX Ten Sequencing System at the J.P. Morgan Healthcare Conference in January. It said “state-of-the art optics and faster chemistry” enabled a 10-fold increase in daily throughput over its earlier machines and made possible the analysis of entire human genomes for just under $1,000.
Plummeting prices should broaden the applications and appeal of such tests, in turn enabling large-scale studies that may someday lead to scientific breakthroughs.
The new sequencers are making their way into the marketplace, with samples now running on a handful of systems that have reached early customers, Chief Executive Jay Flatley said in an interview with Re/code last week. Illumina plans to begin “shipping in volume” during the second quarter, he said.
The Human Genome Project, the international effort to map out the entire sequence of human DNA completed in 2003, cost $2.7 billion. Depending on whose metaphor you pick, the $1,000 price point for lab sequencing is akin to breaking the sound barrier or the four-minute mile — a psychological threshold where expectations and, in this case, economics change.
Specifically, a full genomic workup of a person’s three billion DNA base pairs starts to look relatively affordable even for healthy patients. It offers orders of magnitude more information than the so-called SNPs test provided by companies like 23andMe for $99 or so, which just looks at the approximately 10 million “single-nucleotide polymorphisms” that are different in an individual.
With more data, scientists expect to gain greater insights into the relationship between genetic makeup and observable characteristics — including what genes are implicated in which diseases. Among other things, it should improve our understanding of the influences of DNA that doesn’t directly code proteins (once but no longer thought of as junk DNA) and create new research pathways for treatments and cures.
“The $1,000 genome has been the Holy Grail for scientific research for now over a decade,” Flatley said. “It’s enabled a whole new round of very large-scale discovery to get kicked off.”
Cost breakdown for Illumina’s $1,000 genome:
Reagent* cost per genome — $797
Hardware price — $137**
DNA extraction, sample prep and labor — $55-$65
Total Price = $989-$999
* Starting materials for chemical reactions
** Assumes a four-year depreciation with 116 runs per year, per system. Each run can sequence 16 genomes.
Some have questioned the $1,000 claim, with Nature noting research centers have to buy 10 systems for a minimum of $10 million — and that the math requires including machine depreciation and excluding the cost of lab overhead.
But Flatley defended the figure, saying it’s impossible to add in overhead since it will vary at every research facility.
“Our math was totally transparent and it is exactly the math used by the (National Human Genome Research Institute),” he said. “It’s a fully apples-to-apples comparison to how people have talked historically about the $1,000 genome.”
He also questioned the conclusions of a recent study published in the Journal of the American Medical Association, where researchers at Stanford University Medical Center compared results of adults who underwent next-generation whole genome sequencing by Illumina and Complete Genomics, the Mountain View, Calif., company acquired last year by BGI.
They found insertions or deletions of DNA base pairs only concurred between 53 percent and 59 percent of the time. In addition, depending on the test, 10 percent to 19 percent of inherited disease genes were not sequenced to accepted standards.
“The use of [whole genome sequencing] was associated with incomplete coverage of inherited disease genes, low reproducibility of detection of genetic variation with the highest potential clinical effects, and uncertainty about clinically reportable findings,” the researchers wrote.
Or as co-author Euan Ashley put it to me: “The test needs some tough love to get it to the point where it’s clinical grade.”
Flatley responded that the sample size was small and that the sequencing platforms were several years old. But he did acknowledge they are still grappling with technology limitations.
“What’s hard is to determine whether there’s a base inserted or deleted,” he said. “That’s a bioinformatics problem, not a sequencing problem. That’s a software issue that we and others and the whole world is trying to work on.”
But, he stressed, that shortcoming doesn’t undermine the value of what the tests do read accurately.
“There are many, many, many things where it’s clinically useful today,” he said.
Flatley pointed to several areas where we’re already seeing real-world applications of improving sequencing technology, including cancer treatments targeted to the specific DNA of the tumor rather than the place where it shows up in the body. There are also blood tests under development that can sequence cancer cells, potentially avoiding the need for biopsies, including one from Guardant Health.
Another promising area is noninvasive prenatal testing, which allows expecting parents to screen for genetic defects such as Down syndrome through a blood draw rather than an amniocentesis procedure.
The technology can delineate the DNA from the fetus circulating within the mother’s bloodstream. It’s less invasive and dangerous than amniocentesis, which involves inserting a needle into the amniotic sac and carries a slight risk of miscarriage. Because of that risk it’s generally reserved for high-risk pregnancies, including for women 35 and older.
Illumina, which offers the blood screening for out-of-pocket costs of around $1,500, recently funded a study published in the New England Journal of Medicine that found the so-called cell-free fetal DNA tests produced more accurate results than traditional tests for Down syndrome and Trisomy 18, a more life-threatening condition known as Edwards syndrome.
“It gives some earlier indicators to women in the average risk population if their babies have those problems,” Flatley said. “I think that it will broaden the overall market, and there are other tests that can be added over time.”
But there are ethical issues that arise as prenatal genetic tests become more popular and revealing, including whether parents will one day terminate pregnancies based on intelligence, height, eye color, hair color or minor diseases.
For that reason, Illumnia refuses to disclose those traits that are decipherable in the genome today.
But Flatley said they couldn’t stop purchasers of its machines from doing so, nor competitors like BGI of China (for more on that issue see Michael Specter’s fascinating profile of the company in the New Yorker ). Flatley said there needs to be a public debate on these issues, and he expects that new laws will be put into place establishing commonsense boundaries in the months or years ahead.
“This isn’t something we think we can arbitrate,” he said. “But we won’t be involved directly in delivering [results] that would cross those ethical boundaries.”