4 lessons from a failed IBM-Roche collaboration

The two companies were working together on DNA sequencing technology, but the pharmaceutical giant pulled out Tuesday.

By Forbes Digital Apr 23, 2013 2:57PM
Scientists monitoring computers in control room copyright Martin Barraud, OJO Images, Getty ImagesBy Matthew Herper, Forbes Staff

There's been plenty of great press -- you might even say hype -- about IBM's (IBM) "DNA transistor," a new type of gene sequencing technology that had come out of the company's labs. In 2009 Wired said the device "could revolutionize" genetic testing. Two years later IBM itself picked the device as its "smartest new innovation" for a series published by The Atlantic.  

It wasn't just the media that got excited. In 2010 Roche, the Swiss drug giant, signed a deal to help commercialize the technology.

But Tuesday morning, Roche bailed, announcing that it was ending the IBM collaboration as part of a reorganization of its DNA data business that will also result in the loss of 170 jobs in Penzberg, Germany and Connecticut.

 The DNA transistor, it announced, was not worth pursuing "due to high technical risks involved." IBM hasn't responded to my requests, by email and phone, for more information about the device.

How, in three years, did a tiny device go from deserving sparkling encomiums to getting dropped like a rock? Here are four reasons, which also serve as lessons for thinking about biological innovation.

There really is a first-mover advantage
When Roche and IBM entered into their deal, there was already a company that was dramatically pushing down the cost of DNA sequencing and that had the bulk of the market for the new technology: Illumina (ILMN) of San Diego.

Today, Illumina is more dominant than ever. Yesterday, Illumina announced its first-quarter sales of $330 million, 10% above even bullish estimates. More than that, it said that in the first four months of the year, it sequenced 4,000 human genomes, including 1,000 in a single week. Scientists and even some doctors are using DNA sequencing technology, but they are by and large turning to a single company, with Life Technologies' (LIFE) Ion Torrent division gaining some ground as a potentially lower-cost option. (Life is being purchased by Thermo Fisher/TMO.)

Illumina Chief Executive Jay Flatley likes to say that the DNA sequencing has only shifted to a new player once: when Illumina took it from Applied Biosystems, which made the machines used by the human genome project. Applied had rested on its laurels, and Illumina isn't. Clinical stage DNA-sequencing companies like Sequenom (SQNM), Foundation Medicine, and Genomic Health (GHDX) are using Illumina's machines.

In medicine, there is such a thing as cheap enough
If IBM were the only one developing a smaller, cheaper, chip-sized DNA sequencing device, the technology would still have huge potential. But there are other companies working on technologies similar to IBM's, including Oxford Nanopore of the U.K., which has said it is close to market but has fallen far behind its own deadlines, and Genia of Mountain View, Calif. IBM would have to catch up.

All of these technologies could make DNA sequencing cheaper, faster, easier and more portable -- Oxford Nanopore has shown off mock-ups of DNA sequencers that are literally thumb drives you’d stick in a USB port. But for many applications, Illumina and Ion Torrent are good enough. Since 2007, the cost of the chemicals and machine time to sequence a human genome has fallen from $1 million to about $4,000, and is still dropping. Jonathan Eisen, a geneticist at UC Davis, says that other issues, like preparing the DNA for sequencing or analyzing the data, are more likely to stop experiments from happening than sequencing the DNA itself.

Biotech is not tech
You'd think that when a company like IBM moves into a new field in biology, its fast technical expertise and innovativeness would give it an advantage. Sometimes, maybe, it does: With its supercomputer Watson, IBM actually does seem to be developing a technology that could change the way medicine is practiced, someday. But more often than not the opposite is true. Tech companies like IBM, Microsoft, and Google actually have dismal records of moving into medicine. Biology is simply not like semiconductors or software engineering, even when it involves semiconductors or software engineering.

This is a lesson we've learned over and over again. It's even true for startups that take a Silicon-Valley approach. Take Pacific Biosciences of California, another DNA sequencer maker. The company announced vaporware, raised huge amounts of venture capital, and made an pedal-on-the-gas push to try to take over the DNA sequencing market. It's machine came up short, and the failure to find a niche has left it looking for a purposes.

Beware easy stories
I'm not speaking here of what investors mean when they say "story," but of what reporters like me do. The Web has pushed us all to create more, atomized content. Anything that's easy to write up and sounds cool is more likely to find its way into print. As near as I can tell, this was true of almost every story written on the IBM DNA transistor. There were no deep dives into the creation of the device, or even much reporting. It writes itself: "IBM DNA transistor." Unfortunately, stories that write themselves often don't convey real information.

For the record, I hope Roche is making a terrible decision, and that IBM does introduce a brilliant and wonderful DNA sequencing technology.

More from Forbes


Copyright © 2014 Microsoft. All rights reserved.

Fundamental company data and historical chart data provided by Morningstar Inc. Real-time index quotes and delayed quotes supplied by Morningstar Inc. Quotes delayed by up to 15 minutes, except where indicated otherwise. Fund summary, fund performance and dividend data provided by Morningstar Inc. Analyst recommendations provided by Zacks Investment Research. StockScouter data provided by Verus Analytics. IPO data provided by Hoover's Inc. Index membership data provided by Morningstar Inc.


StockScouter rates stocks from 1 to 10, with 10 being the best, using a system of advanced mathematics to determine a stock's expected risk and return. Ratings are displayed on a bell curve, meaning there will be fewer ratings of 1 and 10 and far more of 4 through 7.

123 rated 1
266 rated 2
485 rated 3
660 rated 4
586 rated 5
652 rated 6
640 rated 7
504 rated 8
289 rated 9
159 rated 10

Top Picks

TAT&T Inc9



Top Stocks provides analysis about the most noteworthy stocks in the market each day, combining some of the best content from around the MSN Money site and the rest of the Web.

Contributors include professional investors and journalists affiliated with MSN Money.

Follow us on Twitter @topstocksmsn.