Cloning Facebook's phenomenal success has not proved easy for scientists eager to develop a network focused on shared research interests. The number of scientific networking websites is growing, but none has emerged as a 'go-to' system.

The indifference could stem from several factors: lack of confidence in individual networks; concerns that personal data might be sold on; or the fact that no one site provides tools or features valuable enough to lure a majority of busy scientists. Despite the proliferation of networks, a big question remains, says Laura James, chief operating officer at the Centre for Applied Research in Educational Technologies at the University of Cambridge, UK. What can they offer that more established sites, such as Facebook, don't?

A number of science networks are attempting to answer that question by offering extra features. ResearchGATE, based in Cambridge, Massachusetts, has launched private fee-based accounts to give institutions or companies their own sub-communities. “Activity levels in the sub-communities are much higher” than in the network at large, says Ijad Madisch, chief executive and co-founder of the website. He says that companies want a communication solution — and ResearchGATE is happy to provide it. According to Madisch, the website has created 30 sub-communities so far, and has a total of 700,000 members around the world, the most yet achieved by any science network, he says. At its launch in May 2008, ResearchGATE, like most such sites, focused on attracting single users — but in future, it will also cater to the needs of sub-communities, says Madisch.

Some scientists may be wary of sharing too much, which probably hinders network adoption. “There is tension between collaboration and competition inherent to science,” says James. “How willing will users be to share advantageous information, even with their closest collaborators, when they are still competing for grants?” Users may also be suspicious of a site's commercial intentions. “How these networks are monetized, and how their user communities will feel about the business models that emerge, remains to be seen,” says James. The creators of Mendeley, a social network that also hopes to attract institutions and companies, have vowed not to sell identifiable personal information.

Networking websites offer an increasing number of features, including virtual conferences, to entice researchers. Credit: LABROOTS/BIOCONFERENCE LIVE

Aspects of some social sites have spiked in popularity — notably, virtual events. Greg Cruikshank, chief executive of the networking site LabRoots, based in Yorba Linda, California, says that his company started BioConference Live, a series of online-only events, last year with much success. The first show, in November 2009, had more than 10,000 attendees, who were given access to 70 sessions at which they could question speakers (in future, Cruikshank expects closer to 40). They could also browse a virtual lobby, exhibit hall and exhibitor booths. “Although LabRoots does a bit of everything — from news feeds to events directory, jobs board and blogs — what distinguishes us is the virtual-event arm of the business, and we'll continue to push that,” says Cruikshank.

Top-down or bottom-up?

The bells and whistles vary, but all the sites have the same aim: enhancing collaboration. To that end, an open-source networking movement is afoot. Rather than creating buzz among users and building a community from the ground up, the creators of freely available research-network software — such as VIVO, developed at Cornell University in Ithaca, New York, with a US$12.2-million grant from the National Institutes of Health, and Harvard Catalyst Profiles, run by the Harvard Clinical and Translational Science Center in Boston — are among those working to forge a network of scientists by wooing whole institutions to adopt their platforms and upload profiles of their faculty members. “Openness is a big deal as VIVO carves out its niche,” says Mike Conlon, the site's director.

This approach is gaining traction. On 9 October, the US Department of Agriculture (USDA) was the first federal agency to join the 14 academic institutions worldwide that contribute to VIVO's 21,000-member network. Harvard Catalyst Profiles has more than 30,000 users, from institutions including Harvard University and the University of California, San Francisco. “We believe that data used in networks need provenance and structure — some kind of hosting institution or scientific society verifying the credibility of data,” says Conlon. Sharon Drumm, a staff officer for the USDA Agricultural Research Service in Beltsville, Maryland, says that the USDA wants VIVO to connect its 3,500 researchers with the hope of fostering collaborations, providing a single information repository for the public and allowing prospective retirees to pass on institutional knowledge.

Conlon cautions that some networking sites simply ingest publicly available data, creating profiles of people who aren't users. Sites with no checks on data run the risk, he says, of offering less valuable information. James suggests that a high enough proportion of fake profiles could undermine entire sites. In principle, the VIVO and Profiles models are more reliable, because they use vetted data.

VIVO hopes to link profiles with not only publications or grants, but also the most important thing that scientists share — data. “Right now data are the stepchildren of this thing,” says Conlon. Linking data sets to people would allow potential collaborators to see what others can offer, he says.

But the existence of a network of profiles, even a vetted one, doesn't necessarily mean that researchers will use it as more than just another database. “I think there's a difference between scientists actively using a network and simply signing up all the scientists in an institution,” says Mendeley chief executive Victor Henning.

It is clear that no single site will meet all scientists' needs. What isn't clear is which combination of sites will. “Scientists are not really interested in social networking as an end in itself,” says Henning. “They network to boost productivity.”