By Dave Brambert
May 7, 2002 | In the battle against AIDS, speed is essential when developing new treatments. That is the driving principle behind the Waterford Project, a comprehensive initiative by leading AIDS researchers to develop a new collaborative approach to more quickly identify HIV/AIDS vaccine candidates for clinical trials.
The project uses off-the-shelf collaborative software applications, existing high-speed campus networks, and the high
|Waterford's Battle for Funding
|For all of its hopes to help fight the war against AIDS, the Waterford Project must first win its own battle for funding or perhaps face shutdown.
bandwidth of the Internet2 nationwide network to enable this new way to do research. Specifically, Waterford taps this existing technology and high-speed infrastructure to enable a multidisciplinary group of geographically dispersed researchers to work together. If funding problems don't derail the initiative (see "Waterford's Battle for Funding,"
) the Waterford Project may spark a new way to approach much of life science research.
"We're entering a new era," says Dr. Warner Greene, who heads the Gladstone Institute of Virology and Immunology at the University of California at San Francisco (UCSF). "In the biological sciences, it is more important who is next to you in your lab than what kind of Internet connection you have. Now we can harness an exciting technology that allows us to bring everything together regardless of geography."
But the project does more than just provide a physical means for researchers to collaborate. The technology lets participants change the basic way they conduct research.
Traditionally, funded research projects include set milestones. If in the course of conducting the research it becomes apparent the work must go in a different direction, the scientist is usually locked into the original track and must complete that work before the more promising work can be started.
The Waterford Project uses the collaborative environment to work around this problem.
Researchers participating in the Waterford Project review each other's efforts, make decisions about research priorities by consensus, and allocate resources based on these priorities. This allows for quick changes in research direction based on results as the work progresses.
Original Thinking The inspiration and initial funding for the Waterford Project came from John Evans.
By anyone's standards, John Evans is an accomplished man. Now chairman and CEO of Evans Telecommunications
| "We're entering a new era. In the biological sciences, it is more important who is next to you in your lab than what kind of Internet connection you have. Now we can harness an exciting technology that allows us to bring everything together regardless of geography."
—Warner Greene, Gladstone Institute of Virology and Immunology at the University of California at San Francisco
Company, he made his first big mark in the communications business by cofounding C-SPAN in 1977. Evans also served as president of Hauser Communications Inc., a cable television company, and later sold the cable properties to SBC Communications Inc. — the first time a regional Bell operating company purchased a U.S. cable system.
His passion to help in the battle against AIDS inspired the Waterford Project. "Every day 8,000 people die of AIDS," he says. "Forty million people are infected. Every day 15,000 people get infected. South Africa did not have AIDS in 1990, yet 25 percent of people there will die from it. In the Botswana mines, 30 percent to 35 percent of the workers are infected. In the United States, AIDS is the No. 1 killer of African-Americans between the ages of 25 and 44."
Evans' interest in the subject goes back to the mid-1980s when he met Dr. Robert Gallo, codiscoverer of the virus now known as HIV and founder of the Institute of Human Virology at the University of Maryland. The two men continued a dialogue about HIV and AIDS for a number of years. It was clear to them that the effort to find an HIV vaccine needed a fresh framework.
Evans' experience in telecommunications led him to conclude that although an AIDS Manhattan Project in which leading researchers were scattered geographically was not feasible, a virtual Manhattan Project was not only possible, but crucial for making fast progress. Highly talented researchers and clinicians were located at universities and research institutes across the United States. If these people could collaborate, it might speed the research process and change the way research is done.
With Gallo, Evans identified the likely and most valuable members of a potential collaboratory (see "Partners in the Collaboratory"). One choice for the complex project was obvious. "We wanted the University of Michigan's help," says Evans, "because we wanted people knowledgeable in electronic collaboration."
Indeed, Michigan's School of Information is the site for the Collaboratory for Research on Electronic Work (CREW), which focuses precisely on the problems and methods involved in a collaboratory. Led by Gary Olson, CREW has studied human/computer interaction, collaborative technology sharing in the form of remote scientific equipment, and technology's impact on human collaboration.
In August 1999, prospective participants met at Evans' Waterford Farm, a Black Angus cattle operation near Middleburg, Va., to start the collaborative process over a number of weekend meetings. "It's one thing to get these high-level scientists to be willing to participate — and that took some doing," says Stephen Effros, president and CEO of the Waterford Project. "It's quite another to get the agreements from all of the institutions." Yet they did get the agreements, including rights to the intellectual property.
Building the Framework Luckily for CREW, the institutions involved in the Waterford Project are located in large cities, with gigapops at the ready. (A gigapop is a gigabit-level regional data transfer center point of presence; it transfers large volumes of data between regional networks and nationwide optical networks.) All Waterford Project participants use regional networks to connect to the optical backbone of Internet2, the big daddy of nonmilitary networks mainly populated by academic institutions.
Specifically, participants connect to Abilene, one of two 2.4Gbps pieces of the Internet2 puzzle. Abilene is a national optical network that was developed in partnership with Qwest, Nortel, and Cisco. Waterford Project participant Harvard AIDS Institute, for example, connects its campus network to the regional Northern Crossroads (NoX) network that serves New England academic institutions. In turn, NoX meets Abilene at the gigapop locations. Each region of the country has its own version of the NoX regional network.
"Typically," says Doug Van Houweling, president and CEO of Internet2, "you'll find that the participants have 100Mbps Ethernet at the desktop, gigabit Ethernet on a campus backbone, or perhaps they use ATM [Asynchronous Transfer Mode] running at OC-12 [622.08 Mbps] or OC-48 [2.488 Gbps]." Abilene will carry all of the videoconferencing and other collaborative traffic between Waterford Project participants.
CREW planned and installed a similar framework for Great Lakes Regional Center for AIDS Research (CFAR) involving Northwestern University, the University of Minnesota, the University of Michigan, and the University of Wisconsin at Madison, a project funded by the National Institutes of Health.
"We've used a lot of commercial, off-the-shelf applications," says Stephanie Teasley, CREW member and senior associate research scientist at the University of Michigan's School of Information. "Principally we use NetMeeting [Microsoft], and the Mac users can get in using Virtual PC for Mac [Connectix].
"CFAR members often have meetings where all the stakeholders in a data set can review samples together," she says. "They can
| "CFAR members often have meetings where all the stakeholders in a data set can review samples together. They can review with a clinician, a person doing raw science, and a pathologist all in real time."
—Stephanie Teasley, University of Michigan's School of Information
review with a clinician, a person doing raw science, and a pathologist all in real time. In one case, the actual patient was part of the NetMeeting session, and that person was thrilled to see their contribution to science in action."
CREW has experimented with other software as well, including ICQ from AOL, WebBoard from Akiva, DocuShare from Xerox, and PlaceWare. "Early on, we tried to deliver video," Teasley says. "We found that most of the scientists closed the 'people' window, and opened the data windows; they wanted to see the data. They already knew each other, so they didn't need to see the faces."
Waterford Project participants already have ViaVideo and ViewStation H.323 products from PolyCom. These systems will be used for videoconferencing among the geographically dispersed participants, from San Francisco to Boston. The Waterford Project will use one-to-many speaking sessions to deliver information, with some one-on-one or group-to-group conferencing. In these cases, seeing the speaker is an advantage, and much of the interaction will be video-based. For a single speaker presenting to the entire participant audience, Teasley says, PlaceWare fits the bill nicely.
One of the only custom pieces to the solution is a database structure that can
|Partners in the Collaboratory
|Waterford Project participants include some of the leading AIDS research scientists and institutions in the country. Each participant has its own research specialty, and each connects through a regional provider to the Internet2's Abilene network.
be shared among all players. The Waterford Project uses HIVWatch, which was developed by UCSF in 1999 to sort treatment outcomes and other information.
By participating in this virtual Manhattan Project, the scientists and advisors make a commitment to do a peer review of research concepts using the technology. The new review process represents an enormous time benefit, reducing the time required to complete peer reviews from months to days.
'Follow Their Nose' The Waterford Project's funding structure does not match grant dollars to milestones as is the norm with most funded research. "We want the scientists to follow their nose, not follow the milestones because the money tells them to do that," says Michael Goldrich, VP and treasurer of the Waterford Project. "A scientist can meet with the board of the Waterford Project, and within a week can go down a new exciting pathway that has been discovered." Goldrich estimates a 10- to 20-fold decrease in the time needed for the review process.
Perhaps the biggest potential benefit of the Waterford Project is its role in assisting development of a new vaccine that actually prevents individuals from getting the virus. Gallo is making progress in this regard by producing antibodies that will exist in mucosal surfaces, the very beachhead of HIV. Such a vaccine has never been developed; even the polio vaccine can't prevent the polio virus, it only manages it. "If we can make it work and sustain production, it will be a huge leap," Greene says.
Such advances take money. Evans estimates that the Waterford Project will need $11 million per year over a five-year period. "Given the scientists involved, we have a great science plan," he says. "This is not my field, but I think if we get these great people to work together, it is the best strategy for finding a vaccine."
Dave Brambert is president of Gilbert Information Systems in Bloomington, Ill., and can be reached at firstname.lastname@example.org.