Novartis shares diabetes genomic data, and experts say there's more to come
By Stephen Pincock | February 26, 2007
Swiss drug maker Novartis this month made the results of a genomic analysis of type 2 diabetes freely available on the Internet. Such open sharing of data might run counter to the general view of the pharmaceutical industry, but many academics see it as part of a growing awareness among firms that there are benefits to be had from making at least some information publicly available.
"Data sharing is good, and it's good to see pharma catching up with academia in this respect," Mark McCarthy from the Oxford Centre for Diabetes, Endocrinology and Metabolism, who is studying diabetes genomics, told The Scientist.
Peter Suber from the Scholarly Publishing and Academic Resources Coalition pointed out in an Email that the European Research Advisory Board (EURAB) recently urged the European Commission to mandate open access for EU-funded research, and half of EURAB members are in industry. "As EURAB shows, industry is starting to join with academics to deliver this message."
The new diabetes data came out of a collaboration between Novartis, the University of Lund in Sweden, and the Broad Institute of MIT and Harvard. The researchers performed a whole genome association study in 3,000 Scandinavian individuals, half of whom with type 2 diabetes. Their aim was to search for genetic variants that influence risk of type 2 diabetes. The results are available on the Broad Web site.
Making the data freely available -- the underlying principle of open access -- had been an important condition of the collaboration, Leif Groop, one of the study's principal investigators from Lund, told The Scientist. "Collaboration between two academic institutions and a drug company could be problematic if we would allow patenting of results," he said via Email. "This was a way in advance to make it possible to work together without having to compete with each other on patent issues."
Furthermore, making the data freely available enables additional labs to follow up on the project, which would take too many years if left in the hands of only a few labs, Groop added. "The magnitude of the results makes it practically impossible for one laboratory to follow up all interesting leads. It is better that all laboratories in the world can contribute."
The company saw other potential benefits to working with publicly funded institutions, who tend to want to release any and all data, said Tom Hughes, Head of Diabetes & Metabolism Research at the Novartis Institutes for Biomedical Research. "First, by collaborating with the Broad and Lund groups, we have experienced first-hand how a world-class academic partnership functions. We also have gained first-hand knowledge regarding the design, analysis, and interpretation of large whole genome association studies." Lastly, he said in an Email to The Scientist, collaborations like this are good for morale among the company's scientists.
Susan Gasser from the Friedrich Miescher Institute in Switzerland, which is part of the Novartis Research Foundation, told The Scientist in an Email that by publicly disseminating its results for diabetes, the company might lure in researchers who identify leads from the data.
"I think [Novartis] figures that if an independent investigator has results that will help cure diabetes, [the investigator] might eventually come back to Novartis to collaborate -- since academic labs cannot do drug development," she said. "Thus sharing basic genomic, transcriptomic, and proteomic information, while unusual, makes good sense -- both business sense and research sense."
Other researchers contacted by The Scientist said they expected more drug companies to make data public as high throughput screening continues to produce enormous datasets. "It is a sign of the times and also the way of the future," said Tom Misteli, Senior Investigator in the Cell Biology of Genomes Group at the National Cancer Institute.
"Making datasets available openly is a reflection of the proverbial truth that four eyes see more than two," he said in an Email. "Making datasets openly available allows everyone to try their favorite tool on a dataset," and more easily compare different datasets. Groop agreed. "Ten years ago everyone was expecting to get the 'big fish' and no one cooperated," he said. "We have matured, drug companies have matured," he said, causing more groups to open up their data.
Last month, Nature published a genome wide association study of SNPs linked to type 2 diabetes risk, which found significant associations for 8 SNPs in 5 loci. "I think [the Novartis group] felt obliged to move on in the hurry with regard of our Nature paper," said last author Philippe Froguel from Imperial College London. He said in an Email that the results from the Novartis study were "moderately useful. Especially because they [published] raw data without any further analysis, which may be very misleading for other scientists."
Oxford Centre's McCarthy said the decision of Novartis and its collaborators not to keep their findings secret has already yielded benefits. He and his team are also tackling the issue of diabetes genomics and have shared data with the Novartis group. "It meant that we have been able to exchange data for months now to the advantage of both sides," he said. "Given that it was funded by a pharmaceutical firm, if they had gone down the secretive route neither side would have benefited."
The SNP consortium, which involves several drug firms and biotechs, is another example of pharmaceuticals making data public, McCarthy noted.
In another similar move, Pfizer and Affymetrix have signed up to a public-private partnership with the NIH called the Genetic Association Information Network (GAIN), which set out to determine the genetic contributions to five common diseases. All data from that collaboration, announced last year, will also be in the public domain.
Open access advocate
Stevan Harnad from the University of Southampton, UK, said the open access movement would be reaching out directly to the R&D industry to consolidate support for open access. "The R&D industry is beginning to recognize the great benefits of [open access]," he said in an Email. "They are a perfectly natural extension of the benefits, to them, of publishing their findings in the first place."
Links within this article
"New genomic tool for diabetes", Broad Institute, February 12, 2007.
Peter Suber, Open Access News
S. Pincock, "UK research to be open access," The Scientist, June 28, 2006.
Whole Genome Scan for Type 2 Diabetes in a Scandinavian Cohort
Thomas E Hughes
R. Sladek, "A genome-wide association study identifies novel risk loci for type 2 diabetes," Nature, February 22, 2007.
Two NIH Initiatives Launch Intensive Efforts to Determine Genetic and Environmental Roots of Common Diseases