a massive pile up of data on servers that is crying out for software that can interpret what it means. Knome, founded in 2007 under the guidance of Harvard genomics ace George Church, initially tried a model in which it did personal genome sequencing and basic analysis for wealthy people. That effort mostly fizzled out, but Knome was able to find a new way forward by offering its software service to researchers who could only get so far by using free open source software, or home-brew programs.
Knome has clearly been on a roll the past year. It isn’t yet profitable, Tolar says, but it grew from 30 employees in September to about 50 at year-end. While DNA sequences can now be had for $4,000 to $5,000 apiece, Knome is offering its interpretation service for an extra $750 to $2,000 per genome, depending on how much support customers want, Tolar says.
The big challenge for Tolar will be to see if Knome can expand beyond its academic customer base and start offering genomic analysis to biotech and pharmaceutical companies. That’s the world Tolar is familiar with, from his past work at NormOxys, CoMentis, and Pfizer. He says he believes that Knome will be in position to help major drug companies in a couple key ways—by helping select patients in clinical trials so that experimental drugs are more likely to succeed, and by helping keep existing marketed drugs from being prescribed to patients who are likely to suffer a rare, dangerous side effect.
Lots of companies have talked up the promise of genomics for improving drug development, with little to show for it. But Knome has already done about a dozen projects for biotech and pharma companies, and expects to have completed about 100 by the end of 2012, Tolar says.
While some pharmaceutical customers are bound to be skeptical, Tolar says he’s convinced the market will be lucrative once Knome proves its value. The market for genomic-interpretation to improve clinical trials, and help improve existing drugs, is probably worth “at least a billion dollars,” he says. Once that market is established, then an even more lucrative market will emerge in “medical genomics,” in which physicians will seek to sequence a patient’s genome and use it as a guide for therapy—which has already happened with certain celebrity patients like Steve Jobs and Christopher Hitchens.
Knome obviously isn’t the only company interested in interpreting genomes—Palo Alto, CA-based Personalis was founded last year on a similar idea, and San Diego-based Cypher Genomics is another relatively new entrant. But Knome is mostly up against the tendency of its customers to lean on open source software first, and only turn to a for-profit vendor when they are stumped and really need help sifting through the information their sequencers have spit out.
Like anything in biotech, the proof will be in the data, and if the data is compelling, it will lead to sales. While biotech companies are tough customers, and on tight R&D budgets, Tolar says he’s convinced they will buy the Knome service if the company can truly deliver what it says it can.
“In biotech, there’s a sweet spot where you run trials that essentially bet $20 million to $60 million and then you turn up the card to see if it worked,” Tolar says. “There’s plenty of money if you can improve the odds. I’d gladly pay $1 million if someone could tell me to focus on the right population for my new drug.”
