Microsoft Aims to Help Scientists Move Past Excel, Make Sense of Gene Data Overload

4/28/09Follow @xconomy

Biologists are living in an era of information overload, and Microsoft says it’s making an effort to help them cut through the noise. The Redmond, WA-based software firm is introducing a new program today called Amalga Life Sciences, designed to help lab scientists make sense of the vast reams of genomic data piling up from experiments, which often come from different software applications and are written in incompatible formats.

Microsoft (NASDAQ: MSFT) is rolling out the program today at the Bio-IT World Conference in Boston. It enlisted Jennifer Chayes, one of the leaders of Microsoft Research in New England, and Jim Karkanias, a senior director of Microsoft Health Solutions, to explain what this is all about to journalists. I got the rundown from Karkanias.

This program is another step in Microsoft’s quest to grab hold of the emerging demand for better information technology in the life sciences industry, a notoriously slow adopter of cutting-edge IT. The company has already been marketing a program for consumers to store their records, called HealthVault, as well as a program called Amalga to hospitals, which it acquired in July 2006. Amalga (pronounced Uh-MAL-guh) is supposed to help hospitals by synchronizing all of the 80 different proprietary IT programs the average U.S. hospital has, to get them to actually talk to each other. Now, instead of just focusing on the delivery piece of the health IT world (hospitals), Microsoft is extending the Amalga concept to basic lab scientists, many of whom are struggling with the same problem of getting their programs to work together well.

Even in an era when sophisticated instruments are making it possible to sequence a human genome with 3 billion chemical units of DNA in each cell, many scientists still try to make do with home-brewed IT systems, and cling to old-school spreadsheets like Excel, which were never intended to do this kind of heavy lifting in computing.

“If you survey the landscape of research, it’s really a patchwork of poorly integrated solutions that people are using,” Karkanias says. “We love Excel, it’s a simple environment for achieving what researchers want to do.” But Microsoft hopes the new program will help integrate the scientists’ existing data with more sophisticated tools to collect it, store it, double-check its accuracy, analyze it, and share it, he says.

Microsoft isn’t saying what this will cost, or how big the potential market is for the company in the broader context of its multi-billion dollar business. It’s not free—it’s an “enterprise software” program for big companies and organizations—and he said it has “quite a few customers already,” although he wouldn’t name names among pharmaceutical or biotech companies. He did say this: “Price has not been a barrier to adoption.”

The first customer that’s willing to be named is just a few miles down the road from Microsoft’s headquarters—the Seattle-based Fred Hutchinson Cancer Research Center. It’s one of the world’s leading biomedical research centers with 2,700 employees and an annual budget of more than $390 million. “We look forward to exploring the potential of Amalga Life Sciences to help us understand data in a rich and efficient way and ultimately help us meet our vision of enabling personalized medicine,” Hutch president Lee Hartwell, a Nobel Laureate, said in a Microsoft statement.

This program is being marketed to major research centers like the Hutch, and to all the Big Pharma companies and Big Biotech companies, Karkanias says. They have different motivations for giving this kind of program a try, he says. The basic researchers see the value in being able to sort through their vast piles of data more efficiently. One analogy comes from cartography. It’s possible to get from New York to Los Angeles by looking at a paper map and following highway turns and directions, but it’s more efficient to let a GPS device navigate along the way.

The pharmaceutical and biotech companies are interested mainly in getting various IT programs to reach efficient conclusions about drug candidates more quickly. The central problem of these industries is that only one out of 10 drugs entering clinical trials ever passes the gauntlet of tests required to win FDA approval, and companies struggle to identify which patients are likely to benefit or suffer unusual side effects, because of their particular genetic profile. They want software to help them increase this success rate, bringing down the cost of development, Karkanias says.

He made an analogy here to the aviation world, in which drug developers are essentially flying now without radar. Once radar came along, it enabled many, many more planes to fly simultaneously in the sky, and to do it much more safely. “Instead of one of 10 drugs succeeding in clinical trials, maybe only one out of 10 will fail,” Karkanias says, of the potential payoff of sophisticated health IT systems. This sounded way too utopian to me, then again, he used to work at Merck, so he’s no greenhorn to the challenges of the pharmaceutical business.

I checked with a hospital IT expert, Tom Wood of Swedish Medical Center in Seattle, to see if the similar version of Amalga is as good as it’s cracked up. He didn’t rave, but the answer was yes. The Amalga version that Microsoft sells to hospitals is easy to use, doesn’t require much support from the IT department, and is good at its stated purpose of fostering compatibility among many proprietary IT programs, says Wood, chief medical information officer at Swedish.

“We liked what we saw,” he says. But Swedish didn’t buy it, because it already has a single integrated software system to begin with (Epic), and no need to iron out the traditional kinks between systems, Wood says.

Microsoft hasn’t yet announced any partnerships with the big companies that make scientific instruments that pump out all this genomic data—companies like Roche, Life Technologies, and Illumina—although Karkanias definitely left me with an impression to stay tuned on that front.

Karkanias also says he’s not aware of any competitors with a product that has the breadth of scope to communicate across IT platforms as Amalga Life Sciences does. Other programs tend to be “highly vertical” or specialized to perform a narrow function, he says.

The big picture is to make drug development less of a risky, conjecture-based business, and provide much smarter predictive modeling of how drugs will behave in the body before spending years of effort and millions of dollars before finding out a drug is a dud. Karkanias used the airplane analogy again, saying that in the early days of aviation, engineers made planes and tested them to see if they’d fly, an expensive and risky proposition. Today, design is computerized, and flight simulation models test a plane long before a prototype is built, which gives those engineers the confidence that it will be assured as safe to fly.

“We want to do that for pharmaceuticals,” Karkanias says.

By posting a comment, you agree to our terms and conditions.

  • joshuadf

    The UW is also piloting Amalga (formerly Azyxxi). I know nothing beyond public information, but the little I’ve seen does not seem to be anywhere near as easy to use as Excel. On the other hand, the researchers I talk to definitely see the need to move beyond Excel.

  • ss

    statistical environment R and scripting language perl do an infinitely better job than excel. anybody serious about informatics in biology uses these or similar tools, not excel. pre-packaged solutions are tricky in biology because of high variance in assay platforms between and even within vendors’ offerings.