By Kevin Davies
December 14, 2010 | Ion Torrent, the semiconductor next-generation sequencing company founded by Jonathan Rothberg that was recently bought by Life Technologies, is offering three community prizes worth a cool $1 million apiece to any research teams that come up with major enhancements in the speed, quantity and accuracy of the company’s Personal Genome Machine (PGM), which is officially released this week.
The prizes announced today are the first three of a total of seven $1-million prizes that Life Technologies will be offering in 2011. Registrations will begin being accepted at the 2011 Advances in Genome Biology and Technology conference next February.
Rothberg, the founder of Curagen, 454 Life Sciences, and RainDance Technologies, unveiled the PGM at the beginning of 2010. The benchtop machine, which costs less than $50,000, detects a pH change each time a nucleotide is incorporated into a DNA strand in one of some 1.4 million features on its prototype chip. The PGM has been tested over the past few months at beta sites including the Broad Institute.
For Rothberg, the cash prizes are all about his dream of democratizing DNA sequencing. “We want to encourage this community,” Rothberg told Bio-IT World. “The ability to win $1- , $2- or $3 million will really get people around the world excited about our technology and working on it. We don’t have a monopoly on good ideas. This allows us to crowd source. Netflix did this successfully for its software to improve selecting films people want to rent. We’re taking the same crowd sourcing concept.”
Rothberg says the decision to sell the company to Life Technologies was to ensure the ability to sell and service the PGM worldwide. “But that’s only half the playbook. If you look at the microprocessor and the PC, it was democratized not just by the company’s actions but by the community. I get tired of [hearing about] ‘3rd-gen’ or ‘next-next gen’ companies. I like defining generations by more than a press announcement, but operationally.”
“Our machine is based on a chip that scales and sequences. The chip is now the machine. More importantly, we want to differentiate it by who uses it and who contributes to it. This generation has to be defined by changing who develops for our platform.”
The PGM costs about $49,000, with each chip retailing for $250. Rothberg says Ion Torrent has “the ability to mass produce, ship them worldwide, and meet any demand, but now let’s get the community developing.
“We want everybody to come out the woodwork. We want all those grad students to say, ‘I’ll buy the machine for $49,000, I’ll run experiments for $250. I might win $1 million. Heck, I might win $2 million if I can win both prizes!’” Rothberg notes, however, that all the prizes will be measured against Ion Torrent’s best in-house performance, “so if people take longer, the bar goes up because we’re getting better internally.” Rothberg says he will use judges at “neutral sites,” possibly large genome centers, to determine the winners.
The prizes fall in three categories: halving the optimal sample prep time for the PGM from about 10 hours to 5; doubling the number of high-quality bases per run from about 40-50 million to 100 million; and doubling the accuracy from 1 error/100 bases to 1 error/200. Rothberg elaborated on each of the three categories:
1) Speed: “If it takes a machine two weeks to sequence [a genome], it doesn’t matter if the sample prep takes 1.5 days. But if you’re getting sequence in two hours, it does!” says Rothberg. This prize will be awarded to a group that can halve the time it takes to go “from a human sample or bacterial or blood to sequence coming off the machine. You could whip it up [perform the sample prep] in ten hours working your butt off! Our first $1 million is for somebody to halve that. We want to help not only the life science labs but also all the translational guys doing clinical research, diagnostics, etc. They want their results faster, and get from sample to data faster.”
2) Quantity: This prize goes to anyone who can extract twice the number of bases as Ion Torrent can on the chip. For all the potential scalability of the Ion Torrent chips, Rothberg acknowledges room for improvement in increasing the number of high-quality sequence reads. “Our first chip, with 1.4 million sensors, produces 100,000 reads of 100+ bases, at less than 1 error/100 bases,” says Rothberg. “We’re just starting, but it’s clear there is a gap between how many great [individual] sequences you get, and how many sensors you have. We’re making huge progress on that gap, but I want the community to help me.”
3) Accuracy: Unlike the first two prizes, the third won’t require access to a machine. Rothberg claims to be able to get up to 40 million high-quality bases per machine (close to 200,000 reads of over 100 base pairs), but instead of an average of 1 error/100 bases, Rothberg is looking for 40 million bases per run at 1 error/200.
Rothberg explains the importance of the algorithms like this: “If we put a huge amount of computer time into a region of the chip, we can double accuracy with better algorithms. But we don’t want the answer to be more computer power, because that doesn’t democratize anything.” Registrants for this prize must use hardware equivalent to the Dell server Ion Torrent ships with the PGM. Raw datasets will be available for download. “The nice thing is, you don’t have to own a PGM, you can just download the files and hack away,” says Rothberg. “I wouldn’t be surprised if an undergrad somewhere in the world won it.”
“Now we have tens of thousands of people out there, and hundreds of thousands of programmers we can democratize this. I’m pretty excited. It solves ‘the Innovator’s Dilemma.’ [Author] Clayton Christianson talks about companies so focused that they miss the next great leap. We want to use the community to solve that. Our chips will get denser and faster beyond anyone’s wildest dreams.”
In detailing the three prizes, Rothberg sounded as exuberant as ever. Some of that may be attributed to the recent birth of his fifth child, Jacob. Rothberg witnessed his son have the Guthrie heel-stick test, in which a small blood sample is taken from the newborn’s heel to screen for a few dozen genetic disorders.
“At that moment,” he said, “I realized every kid who has their heels pricked has to go on an Ion machine in the next couple of years. It really energized me, to get these prizes out there and make this totally ubiquitous.”