As the Dow plummeted toward 8,000 in 2008 and banks couldn't get bailed out fast enough, probably the last thing on everyone's mind was the state of biomedical — particularly genomic — research in the United States. Yet when the American Recovery and Reinvestment Act was signed into law in 2009 to prevent the US economy from bottoming out, investment in basic research figured prominently. Of the $787 billion that ARRA allocated, $18.3 billion was set aside for basic research, and the White House estimates that more than $1 billion of those research dollars are being spent on genomics.
Science funding was in desperate need of help even before the financial crisis came along, says the University of California, Berkeley's Michael Eisen. "In some ways the stimulus allowed science to avoid a financial reckoning that was coming, bad economy or not," he says. "We're still in the shadow of the big explosion of funding under Clinton and then the slowing down under Bush, but the hiring and growth of researchers always lagged behind the money."
"The stimulus has dumped a lot of money into science and it's obviously been good — in my case, it's enabled me to hire somebody," he adds. "I think that its effect on science itself is a little bit harder to characterize in the sense that everybody always has a lot of ideas that are not funded."
So far, more than 14,000 grants have been funded by NIH, and researchers are hard at work building new compute centers, developing social networking sites for researchers, and studying cis-regulatory modules of Drosophila.
"The mission of the stimulus was to stimulate spending and to stimulate job creation. Certainly that is being accomplished," says Clifford Reid, CEO of Complete Genomics. "I think Congress got it right in the sense that there was no opportunity for that money to go into bank accounts and be part of the savings process. … By the structure of the US research establishment, this is money that gets spent on a very timely basis."
"I think it is going to have a very strong economic effect," adds the University of Florida's Michael Conlon.
For now, jobs are more secure and the science stimulus is looking good. But on the horizon looms a dark cloud. In a little less than two years, all ARRA funding will have been exhausted. "From a scientific perspective, it's fantastic and whether one ties it to economic stimulus, it sure as hell is a scientific stimulus and that's obviously what's most important to us. It's obviously helped us — and I'm sure others — either save jobs or hire people who otherwise wouldn't have been hired," say Duke University's Hunt Willard. "The hope, of course, is also a fear: how are we to come back to Earth in a few years? Because we don't suddenly want to be laying people off in two years because all of these grants can't be renewed."
All those grants
ARRA set off a flurry of grant application submissions from PIs. More than 30,000 applications were turned in to NIH, and by March of this year, NIH had spent about $4.6 billion of the $8.2 billion it allocated for scientific research. The rest of the NIH funds were spread among lab construction, facilities expenses, shared instrumentation programs, and other projects.
To get all those funded grants out the door as quickly as possible — the act specified a hard two-year deadline for the funds to be spent — NIH took a variety of approaches, including lowering the paylines and establishing the Challenge Grant and Grand Opportunities programs.
"You really had to have an idea ... that was ready to fly and generally one that had not gotten funded before or that you hadn't bothered to apply for funding on," Berkeley's Eisen says. His lab received $184,737 to model cis-regulatory modules in Drosophila. He adds that the short turn-around time allowed for some interesting projects to get funded, and that some of the constraints typically encountered in the review process were relaxed.
[ pagebreak ]
One of the tenets of the ARRA program was that it would fund "shovel-ready" projects. While that may have brought to mind people in hardhats working on beat-up roads in California or North Carolina, these were projects ready to break ground in genomics. Washington University in St. Louis received $14.2 million to expand the genome data center at its medical school. Elaine Mardis, co-director of the genome sequencing center there, says that with all the next-gen sequencing data they're generating, the center was rapidly outgrowing its capacity. This was, however, something they had predicted when designing the original center. "It was built for expansion where we wouldn't have to duplicate the infrastructure — which means all of the redundant power, back-up chilled water, all of the humidity controls, and the things that go into the data center environment maintenance," Mardis says. "All we really needed to do was build out another approximately 4,000 or 5,000 square feet for additional racks and storage."
The ARRA funding came along at an opportune time for WashU to increase its capacity. "The ARRA funding to us seemed like a tremendous opportunity to try to qualify for the funding that was available for construction and use to essentially to build out another data floor," she adds. They plan to break ground on the center this month and hope to have it completed by the fall. The new computing center will allow researchers at WashU to soldier on with current cancer genome sequencing and analysis plans as well as to tackle new microbiome samples, Mardis says.
Similarly, at Duke, Willard's Institute for Genome Sciences and Policy received $461,402 to build a high-performance computing system. He says that this equipment will speed up a lot of analysis at Duke. This idea had already been in the pipeline, he says, but ARRA presented a good opportunity to make it a reality. Willard says that when he started at the institute more than seven years ago, he had begun thinking about building a shared cluster compute resource to be housed there, and that would also be "used broadly across the campus by even people who wouldn't know a genome if they bumped into it."
IGSP also invested heavily in the Encode and ModEncode projects. "The data sets that were being generated because of that and sequencing data sets were just getting enormous," he says. "We were ready for a quantum jump, and this grant came along at the right time to allow us to do that."
In addition to building out centers, stimulus money is also going to shared instrumentation, and bringing more people into next-gen sequencing. "The biggest impact from my perspective … is going to come from the shared instrumentation grant program that NCRR funds," says Mark Lively at Wake Forest University School of Medicine. "That program funds both intermediately expensive instruments and what are called high-end instruments. A significant number of the requests — there were more than 1,200 requests for shared instruments — were for next-generation DNA sequencers."
The funding is not only affecting academics, but also sequencing companies. "Absolutely, the stimulus has had, I think, a tremendous impact on the appetite for sequencing technologies generally, and for ours specifically," says Pacific Biosciences' Steve Turner, who adds that a good portion of the company's customers have been ARRA-funded.
[ pagebreak ]
PacBio itself received ARRA grants. One grant is a continuation of a technology development grant that it will use to study ways to extend and document its single-molecule, real-time DNA sequencing. The second grant is to develop an application of its technology for epigenetics. "We can see epigenetic modifications to DNA without any kind of upstream conversion through bisulfite because of the fact that those base modifications have an impact on the kinetics of DNA incorporation. Because we are looking at the system in real-time, we are sensitive to all those kinetics," Turner says.
Complete Genomics' Reid adds that at least two of their customers are ARRA-funded, and have initiated projects with the company worth more than $1 million. That, he says, has helped Complete grow and hire more people.
More generally, Reid says that the stimulus money is reaching companies in the life science space. He says that his competitors have said that stimulus money has shown up more slowly than originally expected, though it has reached them. "It's had a really material effect on their financial results and that's been, I'm sure, good for jobs ... as well as good for the advancement of research in the area of life sciences," Reid says.
Another large-scale project aims to foster collaboration. Building off a system developed at Cornell University, Florida's Conlon is building a nation-wide semantic Web system, called VIVO, with $12 million in ARRA funding. VIVO compiles the research interests of faculty members. "[There are] algorithms that will go through publication records, grants, and other background data and say, 'Gosh, we've looked at your profile data and you're interested in X, Y, and Z.' That's nice but the faculty member is the ultimate authority on that," Conlon says, adding that the faculty members have control over their information.
The algorithms are built off of MeSH terms from the National Library of Medicine, which allows Conlon to infer more details about researchers' interests. "Somebody can say, 'I'm interested in human papillomavirus' [and] if someone says that, we would be able to infer — and infer with some accuracy — that they are interested in human causes of cancer," he says. "They didn't have to tell us they are interested in human causes of cancer. We know that with some precision because of their use of a MeSH term."
"I think what we're doing is really creating the foundation for the future of team science," Conlon adds.
Measuring the field
Even with all this money flying around, it's difficult to determine whether the stimulus has been successful in aiding science — and it may be too early to tell. "We're not going to be able to point to any one breakthrough or discovery or cure or advantage to health that is a direct result of the ARRA money until we look back at least five years from now," Wake Forest's Lively says. "In part, that view is going to be confused or impacted by the ability or inability of NIH to sustain those projects in FY11 or FY12 after ARRA is gone."
The influx of funds into biomedical research has reinvigorated the field, bringing it out of a long fog, many say. A year and a half ago, Complete Genomics' Reid says, the general feeling of the field was bleak. "There was a real, palpable depression in the life science community at what turned out to be the beginning of the financial crisis," he says, adding that researchers were actively discussing how to cut back.
"When the stimulus came along, all of a sudden, the growth [of the NIH budget] was restored," he says. "The entire tenor of the community changed and it changed within 90 days to be able to see that NIH was back on a growth path."
[ pagebreak ]
WashU's Mardis agrees. "Overall, it's brought a lot of enthusiasm back, at least to the biological sciences, to have these additional monies available," she says.
Economically, though, the impact is a bit grayer. Science, of course, can have an economic impact — the case in point there is Google, says Julia Lane, a labor economist at the National Science Foundation. Google co-founder Sergey Brin was a postdoc when he received a grant that allowed him to develop the Markoff processors that form the basis of the search engine.
In the short term, among the metrics being logged is the number of jobs that have been saved or even created. Each ARRA grant recipient must submit quarterly reports that include who is receiving the stimulus funds, the status of the project, and the number of jobs created. Robert Birgeneau, chancellor of the University of California, Berkeley, says his school has been able to save or create nearly 150 jobs. Lively, who is also the president of the Federation of American Societies for Experimental Biology, says that a good chunk of his salary from Wake Forest is derived from ARRA funding.
"On the economics, the question is job creation. It's the health of the institutions that are housing a lot of the biomedical research, and restoring health to institutions — universities, and colleges especially — which are still suffering a lot and whose recovery is lagging from the financial downturn, probably lagging a couple of years behind the rest of the economy," adds Duke's Willard.
There's a nonlinear relationship, Lane adds, between scientific investment and economic growth. To make better sense of the effect of investing in the sciences, a team led by Lane has developed STAR METRICS — for Science and Technology in America's Reinvestment: Measuring the Effect of Research on Innovation, Competitiveness, and Science. For this project, she is partnering with universities and PIs to develop infrastructure to collect data and other accounting measures, including jobs created, citations, patents, and other factors, in an automated way. "The basic idea here is to, instead of trying to make the poor researchers respond, just automate the capture of the information. A big concern is that about 40 percent of a PI's time is spent just in reporting and that's a real competitiveness issue. If we can automate DNA sequencing, for crying out loud, we should be able to automate PI reporting. So that's what we're focusing on," she says.
Lively also suggests looking at infrastructure. "I think the metrics that I would use on a shorter term would be to look at the building program. NCRR is still in the process of awarding about a billion dollars in ARRA money for innovations and modifications to existing labs and facilities," he says. "I think in three years', four years' time those projects are completed and occupied. You would be able to go and look at those institutions and really assess the impact that's had [on institutional research]."
Over a longer term, Willard also suggests that an eye be kept on human capital and their funding. "Have we restored the growth of the NIH budget? Have we restored a group of young investigators — whether they be students or just faculty starting their careers — so is the pipeline full of the next generation of scientists as people like me get off the stage?" he asks. "I would take a real deep look at that pipeline."
However, Berkeley's Eisen notes that narrowly constrained metrics don't often give the full picture of science's impact. "It's a big challenge. How do you measure success of a research investment? It's an important question, but very difficult," he says. "The number of papers is a poor measure of the importance of a research field. Some fields publish a lot of papers. ... I'm just uncomfortable thinking that narrowly about science."
[ pagebreak ]
Ironically, he points out, science doesn't lend itself well to being measured. "The question of how important the stimulus effect has been is: has it made the people who got funds better? Have they done better science? And secondly, can you measure somehow the knock-on effects of that to make everybody a better scientist? You need a metric for figuring out how effective individual scientists are," he says. "But the ones we have aren't just inexact; they are terrible."
After the boom
The stimulus grants, though, are of finite funds and for a finite time — none lasts more than two years. While this sudden influx of funds has brightened the mood in the field and saved jobs, the question is: what happens when that influx stops? Will a bust follow the boom?
"I don't know," NSF's Lane says, adding: "There you go. You've got an economist telling you, 'We don't know.'"
Eisen traces science's funding crisis back to the oscillation of fund-rich and fund-poor times. During the Clinton administration, there was an expansion of NIH funding. According to FASEB, between 1995 and 2000, the NIH budget climbed from $1.9 billion to $2.5 billion, in today's dollars. That expansion was then pulled back during the Bush administration when the NIH budget topped out at $35.3 billion in 2004, before falling to $31.5 billion in 2008, and not keeping pace with the anticipated 9 percent annual growth for NIH's budget. Now, the stimulus is providing another explosion of funds — and another anticipated drop-off.
Eisen likens the situation to a wildlife ecosystem that's not being well managed. If you keep putting food into the system, the population of deer (or your animal of choice) will grow, and if the food is removed, the deer suffer. "The stimulus stuff scares me in that sense. It's essentially two completely unmanaged events: one is this great infusion of money and the second is this extraction, removal, of all that money," he says. "I don't think that anybody's really thought about how that is going to affect things, even beyond the immediate there-will-be-fewer-people-with-jobs stuff."
One potential way to soften the blow of stimulus funding withdrawal is through better funding of the science agencies in 2011-12. That cannot, of course, be enough money to continue all the ARRA grants. "ARRA basically brought us back to the curve that we would have been on if, post-doubling, we would have gotten a sustained and predictable funding since 2003," Lively says. "ARRA's really bumped us right back onto that curve and we're going to fall off of it again."
"I trust in Francis," adds Duke's Willard. "I think he has been — and is going to be — a very strong advocate for upward pressure on the budget, and he obviously has a receptive White House and Congress. I think that at one level, we'll do just fine and we'll creep up slowly a few percent of the time and that'll be a good thing."
In his budget proposal for 2011 — which keeps discretionary spending flat — President Obama suggested a 3.2 percent increase in the NIH budget for 2011, which would increase the grand total by $1 billion to $32.9 billion. This, FASEB's Lively says, shows that the current administration is taking science seriously, though FASEB would like to see a larger gain. Its recommendation for 2011 is $37 billion. Lively says that number was decided on after considering the true allocation for 2010, which for NIH was around $35 billion — that number includes the 2010 portion of ARRA. Then FASEB added about a 3 percent increase. "It is $5 or 6 billion above the president's FY11 recommendation, which is, by the way, the highest recommendation we have received in more than eight years. It's very generous, particularly in the face of an attempt to freeze the discretionary spending," Lively says. "But $37 billion is what we — and we understand some folks within NIH — have suggested as necessary to sustain the current level of effort."
[ pagebreak ]
"The bottom line is that the federal funding has to at least take into account some reasonable adjustment. You don't want to go from the top of the mountain to the bottom of the valley — the bottom of the valley representing the previous level of funding for the NIH," WashU's Mardis says. "It would be, I think in some ways, fairly catastrophic for the ARRA funding to expire, and then for us to be back where we were before."
One of the groups that is most vulnerable to the ups and downs of science funding is young investigators. As funding tightens, the percentage of funded grants drops — the average age for a first R01 is 42 years old. As Eisen points out, the pool of researchers swells to receive the money and when it dries up, they'll be out of luck.
Willard would also like to see a focus on training in upcoming NIH budgets. "I am very concerned that the pool of money available for training grants and really creative interdisciplinary kinds of training is just not there. There's a large effort that's gone into K awards, usually for taking people who are clinically trained and sort of getting them into the science mill. It's not that I'm against that but I think there is a very important gap to be filled at the interface between life sciences research and quantitative research," he says. "The genomics crowd ought to be driving that — we're not the only area of science where that's true, but we ought to be driving that because it is so self-evident to all of us. Our students now are not just experimentalists and not just computationalists."
Complete's Reid and PacBio's Turner also suggest that the private sector may step in to fill some of the void left by the end of ARRA grants. "As the stimulus spending winds down, I think Congress hopes and expects that the private sector will pick up the slack," Reid says. "The timing of that is impossible to call exactly, but I think we are already through the worst of this financial crisis, and so the stimulus money rolling out now has accomplished exactly its mission: not to replace private sector activity, but to buffer, to cushion what otherwise could have been a quite devastating crash."
Even if in five or 10 years researchers look back at the stimulus with better metrics in hand, they still might not be able to say with any definitiveness what the impact of ARRA was on science and the economy. As Eisen points out, it wasn't a controlled exercise. "The stimulus is an interesting experiment at many levels; it's an interesting experiment economically, it's an interesting experiment on the scientific level. The same things that plague the ability to analyze its effects on the economy plague its ability to analyze its effects on science, namely that it's not a controlled experiment," he says. "We don't know what would have happened had this money not gone into the economy or science. We really have to guess and that's not something I think we're good at. It's really hard to know what its impact on science is and at some deep level, it will probably have to remain ambiguous forever."