Complete Genomics Stuns Genomics Community, Confirms Industry Shattering Capability

1 1 Loading

Daniel MacArthur at Genetic Future has just completed his post covering the Advances in Genome Biology and Technology (AGBT) conference, where an intense battle is taking place as genetic sequencing companies seek to position themselves in the race to bring fast, cheap genetic sequencing to the masses.  The verdict: Complete Genomics has stunned the genomics community by confirming it is on track to blow past the competition.

When reading Daniel’s post you can feel the intensity at the conference as companies at AGBT fight for top bragging rights to deliver the fastest and cheapest genome sequencing capability to the world.  The stakes are extremely high: literally billions of dollars in sales await the company that can dominate in the delivery of affordable, rapid, whole genome sequencing to every human on the planet.

The intensity at the conference reached a climax when Clifford Reid, the CEO of Complete Genomics, delivered his much anticipated presentation to confirm whether or not the company’s industry shattering claim of $5000 sequencing of a whole human genome by mid-2009 was really true.  According to MacArthur, Complete Genomics did not disappoint, and hence a company that nobody had even heard of until it came out of stealth mode in October 2008 now appears to have cemented its position solidly at the front of this high stakes race.

So what is the secret behind Complete Genomics’ stunning success?  The secret is that Complete Genomics’ has completely upended the business model for genetic sequencing, focusing on computational might and centralization while its competitors have focused on flexibility and distribution.

Complete Genomics’ competitors have built their business models around the production of devices that their clients can purchase and install at their own facilities.  This business model offers customers the flexibility of performing their own genetic sequencing, but in exchange for this flexibility customers are limiting the computing capacity they can throw at the problem to what can be squeezed into a deliverable product.

Complete Genomics, on the other hand, does not produce anything that its customers can install or use.  Realizing that genetic sequencing is a raw horsepower problem of who has the most computing capacity,  Complete Genomics has built a server farm that can be continuously expanded to throw more and more computing muscle at the problem.  Complete Genomics has optimized the computation of genetic sequencing, building one centralized, giant super sequencer that now seems poised to leave its competitors in the dust.  Within the next 5 years Complete Genomics is aiming to sequence a whopping 1 million human genomes!

It is important to point out that being the fastest doesn’t mean anything if your output is littered with errors.  MacArthur notes that investigation is still required to confirm the accuracy of Complete Genomics’ sequencing technique when he writes:

“I was convinced by the SNP data, but I will be very interested to see how the system performs in terms of calling large-scale structural variants, and in dealing with highly repetitive regions. These are major problems for very short read technologies that can’t be solved by simply increasing coverage; information on these issues was largely missing from Reid’s presentation, aside from an almost cursory mention of a technology called “long fragment reads” that might help to address such problems. Large-scale structural variants play an important role in human variation and disease, so Complete will need to deal with these areas effectively if it is to generate genome sequences that can realistically be called “complete”.”

As an aside, many might mistakenly think that genomes are already being sequenced for less than $1,000 by companies like 23andme and decodeme, but these companies do not offer full genome sequencing.   Instead they only analyze a few hundred hot spots in your dna called SNP’s that can tell you lots of interesting things about your dna, but not the whole story.  Fully sequencing every single one of the approximately 3 billion base pairs of your dna is a completely different scenario, and this is where the real battle is being fought.

Update: More information has recently been revealed about Complete Genomics

Discussion — One Response

  • dirk bruere February 6, 2009 on 1:35 pm

    This is really going to speed up the discovery of genes associated with everything from intelligence (however measured) to longevity. As an aside, just how much computing power is required? Is it FLOPS limited or I/O limited?