Sunday, July 20, 2008

A Case Study In How The Quest For Funding Tarnishes Science

I must add this article now and preserve and share it. It has relevance to the on-going debate surrounding the myth of man-caused global warming, how it came to be, and why it persists. The fact that the scandal described in the following article originated at my alma mater is an aside. This issue affects all scientists and all of science. I welcome comments. (source)
Peter


Can of Worms

The UW’s worst scientific scandal in decades exposes the ways that external funding has increased the vulnerability of two of the university’s most valuable assets — its integrity and its graduate students.
By John Allen

On September 20, 2006, Irwin Goldman received an e-mail with good news — “exceptional news,” he wrote at the time — he’d never expected to hear, at least not from anyone at UW-Madison. Graduate student Amy Hubert PhDx’08 had made a breakthrough in her research, successfully identifying the piece of DNA in the nematode worm Caenorhabditis elegans that corresponds with a gene called laf-1.Laf-1 is important in sex determination among nematodes, which are typically hermaphroditic. Those that have a single laf-1 gene mutation develop as females. Those that have two die — thus “laf,” which stands for “lethal and feminizing.”No one ever said a worm’s life is easy.But, then, neither is a graduate student’s, and Amy Hubert’s has been harder than most.

Seven months earlier, she was working and studying in the lab of Elizabeth Goodwin, then a leading researcher in C. elegans genetics, when Goodwin was accused of scientific misconduct — essentially of misrepresenting her work in an application for federal funding. It was one of the most serious scandals in UW science, and even today, those connected to it are reluctant to talk about it. The university launched an investigation, and Goodwin soon resigned. UW-Madison had to give up thousands of grant dollars and still faces the possibility of sanction by the federal government.

The chief evidence against Goodwin came from Hubert and her graduate student colleagues, and the aftermath of the scandal has taken nearly as heavy a toll on them as it did on Goodwin. Of the seven people working in her lab — six graduate students and one research specialist — five are no longer at UW-Madison. Hubert hopes to complete her doctorate soon, which would make her the first member of the lab to do so.Her success may illustrate her academic skill, but the fact that she’s the only one from that team nearing graduation to date shows something else: the vulnerable position in which graduate students find themselves in the modern scientific world of grant-based research. Wholly dependent upon their faculty mentors for educational, financial, and career support, grad students lead a perilous life. And the increasingly competitive struggle among professors to win research grants only makes that peril more acute.

Chasing Phantoms
C. elegans is a model organism, one of a handful of species that scientists study closely to learn about general biological principles. There are a few others: several bacteria and yeasts, Arabidopsis plants, fruit flies, zebra fish, rats, and mice. Nematodes are useful as a model organism because they’re among the simplest creatures to possess a nervous system, they can be frozen and thawed without damaging their viability, and they reproduce rapidly — even more quickly than fruit flies. Nobel Prize-winner Sydney Brenner pioneered studies of C. elegans in the 1970s, making it a relatively recent entry into the roster of model species. As a result, says CALS associate dean Irwin Goldman PhD’91, “the C. elegans community is still relatively small and tight-knit.”That closeness is highlighted at WormBase, an online database that keeps records of the researchers who are working on C. elegans, cross-referencing mentors to protégés. Sydney Brenner, for instance, is listed as WB (WormBase) Person77. He was the mentor for David Hirsh, WBPerson259, a professor at Columbia who mentored Judith Kimble, WBPerson320. A biochemistry professor at UW-Madison, Kimble supervised Elizabeth Goodwin, WBPerson213, when she was a postdoctoral fellow. And Goodwin was the mentor to WBPerson3474, Amy Hubert.Hubert came to UW-Madison in 2001 from the north-central Kansas town of Concordia. Like many graduate students, she’d earned her previous degree elsewhere — a bachelor’s at the University of Kansas. She’d been attracted to the UW by its reputation as a leader in genetics, but she didn’t join Goodwin’s lab until spring 2002.“I liked Betsy [Goodwin] as an adviser,” Hubert says. “She was very friendly, very hands-on, always checking in with us about how our work was going.”

Goodwin was then one of the rising stars of the UW’s genetics department. “She was an extremely good citizen,” says Phil Anderson, another C. elegans expert (WBPerson21) and a professor of genetics. “She did more than her fair share of committee work, and she was very involved socially. She entertained prospective graduate students and helped recruit new faculty. She brought a genuine sense of joy to working here.”Goodwin’s lab was then one of only two or three in the world that were doing such advanced research into sex determination among nematodes, aiming to shed light on the importance of sexual reproduction in evolution. She was developing a hypothesis that laf-1 affected sex determination in C. elegans because it encoded a strand of “small RNA” — ribonucleic acid — instead of a conventional protein, which most genes encode. Such a discovery would have been a major coup, had she been able to prove it.“Within the last five or ten years,” says Anderson, “small RNAs have become very fashionable in research. There’s been a lot of exciting progress around small RNAs, and there’s a lot of mystery there.”But Hubert quickly lost conviction that Goodwin’s small RNA hypothesis was the right track. “When I first joined the lab,” she says, “I was getting the same results that Betsy and the others were. But I began to see that the laf-1 mutants weren’t all matching the prediction. Some of the results were just — well, they weren’t all that strong.”

While Goodwin continued to pursue the small RNA hypothesis, Hubert turned to more traditional approaches. She decided to revisit the question of laf-1’s identity by continuing to map its location in the C. elegans genome. “And the more mapping I did, the more it showed me that laf-1 was in a different spot from where everyone else [in the lab] was looking.”“Amy’s a real geneticist, while the rest of us were molecular biologists,” says one of her former lab partners, Mary Allen MS’06. “She was willing to take a long time and go back through the laf-1 work from the beginning, redoing the experiments and looking for inconsistencies. She was willing to go through the tedious process of looking at each generation of worms to see the phenotypes. I can remember her spending up to twelve hours in the lab, even on Saturdays. It was a lot of hard work, and a lot of time, but it paid off for her.”It was this difference of opinion that ultimately saved Hubert’s career at the UW. When the Goodwin affair blew up at the beginning of 2006, Hubert’s colleagues, whose work had all been intertwined with Goodwin’s search for small RNA, saw their years of labor vanish.“The project that Betsy was so committed to turned out to be a phantom,” says Anderson. “Laf-1 does not encode small RNA. And because that was a phantom, it wasn’t producing results, and it never would. The students who were chasing this phantom, they weren’t making any progress. Amy had the foresight — almost from when she first joined the lab — she had the good sense to challenge Betsy on this. She went back to the tried-and-true methods of mapping the DNA. Because of that, she was able to survive Betsy’s departure. That’s why she was able to make her breakthrough.”

Trying to Do Good
The Goodwin affair became public on February 23, 2006, when Goodwin resigned from the UW, bringing to a close three months of tension and quiet recriminations that had infected the members of her lab. She slipped quietly out of town and hasn’t spoken publicly about the allegations since, leaving her lawyers to respond to the charges. But her departure didn’t bring the affair to a close. In many ways, the difficulties were just beginning.The misconduct allegations leveled against Goodwin were, according to Bill Mellon, associate dean of the graduate school and head of the UW’s research compliance, among the most serious he’s heard in his twenty-nine years at UW-Madison, and certainly the most serious involving a tenure-track faculty member. But the university couldn’t just drop the issue upon Goodwin’s departure.“We had an obligation to investigate the charges,” says Mellon. “There were at least two federal grant applications involved, a renewal and an application for new funding, and we have to show the government that we’re serious about our honesty and the honesty of our scientists. So we had to carry out a long and detailed process — and I must say, Professor Goodwin’s absence made the process all the more difficult.”

Without Goodwin to answer the charges or explain her actions, investigators have struggled to determine just how far the damage goes. Mellon is confident that Goodwin wasn’t guilty of fraud — “she used the federal money to run the experiments she said she was going to run,” he says — and so far, no one has challenged the validity of the articles Goodwin has published. But the university had to report its findings to the federal government, and it hasn’t yet heard what the government’s Office of Research Integrity has to say about the Goodwin case.Still, if the process has been difficult for the university, it’s been even harder on Goodwin’s former students.At the time of the allegations, Goodwin’s lab employed six graduate students — Hubert, Allen, Garett Padilla MS’06, Chantal Ly MS’06, Sarah La Martina ’00, MS’06, and Jacqueline Baca — and one research specialist, Maya Fuerstenau ’04. In November 2005, Padilla discovered a discrepancy in the data that Goodwin was reporting in her grant renewal application. The application made it appear that the lab was producing favorable results, when in fact the data were incomplete, inconclusive, or worse. He brought this to the attention of his colleagues and to Goodwin.“At first, all we saw was that the data were wrong in one figure on one of the grants,” says Hubert. “After Garett talked to Betsy about it, she told him it was just a mistake, that the data had been placeholders. But the more we looked, the more we saw that seemed wrong. She was saying some experiments had been done that hadn’t been, and the figures that were being used to illustrate her results weren’t the correct figures. The conclusions she gave, in some cases, weren’t correct. Things had been relabeled, and we concluded that it couldn’t have just been a mistake. It had to be deliberate.”

The graduate students then faced a difficult decision — whether to report Goodwin to the university, knowing that, if they were wrong, it would ruin their relationship with their mentor and boss, and if they were right, it would destroy the research they’d been conducting for years.“We had a lot of discussion about what to do,” says Allen. “A lot of discussion. We were grappling with a big decision, and some people were afraid that they’d have to start over.”Eventually, however, the students came to a unanimous decision. “When we looked at it, though, there wasn’t really a choice,” says Hubert. “We wanted to believe [Goodwin]. We felt she was trying to do good for the lab. But we had to report it.”In December, they informed Mike Culbertson, chair of the genetics department. He reported the allegations to Mellon, and Mellon asked Irwin Goldman from CALS and Paul DeLuca, an associate dean of the School of Medicine and Public Health, to begin an informal inquiry to determine if there was any merit to the students’ allegations.

“We met with the students and with the professor, and we looked at the grant proposal and the data provided by the students, and it seemed clear that there was some attempt at deception,” says Goldman. “Betsy was giving one set of numbers to NIH [National Institutes of Health] and another to the students. This was clearly not a frivolous accusation.”While Goldman and DeLuca were gathering information, the lab environment was deteriorating. “We didn’t really know what was going on,” says Hubert. “We had spoken with the investigators, but were asked not to speak about Betsy to anyone else. At the same time, she was free to defend herself, and we had the feeling that she was bad-mouthing us around the department, saying that we were making mountains out of molehills and that we were out to get her. For a while, we got the feeling that we were the ones who were in the wrong.”As the situation progressed, the students felt increasingly betrayed and isolated. “It’s like there were two Betsys,” says Hubert. “The one who was my friend and who was so helpful, and then the one who had done this thing on her application. I couldn’t believe they were the same person. I still can’t.”

Goldman and DeLuca issued their recommendation to Mellon, reporting their opinion that the students’ charges had merit. Mellon then referred the matter to a formal investigative committee. And just as they began their work, Goodwin resigned. The committee finished its work a month later, concluding in a report to the Office of the Chancellor that Goodwin was guilty of the charges the students had leveled. The UW submitted its report to the Department of Health and Human Services, which oversees NIH, and returned what remained of her grant money.Within a few weeks, news about the misconduct spread through both the local and the scientific community, as major stories appeared in the Wisconsin State Journal and in the journal Science.“Once the report was delivered, I think it was clear that we’d been in the right,” Hubert says.But if they were vindicated, they were hardly safe. In reporting their mentor’s misconduct, they’d endangered their own careers as well.

Heroes
For Irwin Goldman, the key to the Goodwin affair isn’t ultimately what happens to Goodwin, but rather what happens to her former students. Hubert and her colleagues, he says, “displayed professional heroism. They put their careers at great personal risk for the sake of scientific integrity. That’s exactly the kind of people we want to attract into the sciences. These students risked a lot for the truth. They didn’t deserve to suffer for the actions of their mentor.”And yet when the students decided to come forward, it seemed likely that they would be punished nearly as severely as Goodwin. When she resigned and the university forfeited her funding, her lab closed down. That meant the students were out of work — no salaries, no tuition reimbursement, and perhaps worst of all, no research. That would mean no progress on their degrees, and possibly starting over. That, Goldman feels, is a poor reward for honesty.Unlike undergrads, graduate students are often more like apprentice academics, breaking into careers in science and research. Many parts of the university, including the College of Agricultural and Life Sciences, where Goldman works, are dependent upon them for carrying out research. Protecting these students from dishonest faculty isn’t just an issue of justice, he maintains; it’s a matter of the UW’s self-interest.“We have almost a thousand graduate students in this college alone,” he says. “We try to get the best students to want to spend five or more years here. And if we want to attract and keep them, we have to care about them and their vulnerability.”

So Goldman convinced his colleagues within CALS to guarantee that Goodwin’s students would receive funding for at least the remainder of the semester, and he began looking for avenues through which they could continue their studies. For Hubert, who was far advanced in her work, he found a home in the lab of Phil Anderson. Another student, Jacqueline Baca, who was nearer the beginning of her work, was also placed in a different lab. But the others, facing the realization that they would have to start over, left the UW. Allen left Madison to study worms under Tom Blumenthal at the University of Colorado-Boulder; Ly and La Martina found laboratory jobs in other states; and Padilla departed for law school.“We can’t give [students] back the time that they lose,” says Bill Mellon, “but we can try to protect them from financial disaster.”That’s what Goldman set out to do. After Goodwin resigned, he began work drafting a policy for CALS that would protect graduate students in the event of scientific misconduct by their mentors. It demands that the college “use its best effort to secure funding for CALS graduate students, including Research Assistants, Fellows, and Trainees, and research associates whose positions and funding may be jeopardized by his/her good faith disclosure of scientific misconduct.” Adopted last year, the policy “is the only one I know of in the country,” says Goldman, and it’s since become a model for a university-wide attempt to improve the grad-student safety net.

“If there’s a hero in all this, it’s Irwin,” says Anderson. “He not only did everything he could to do right by Betsy’s students, but he also pushed the rest of us to try to do right by them.”Still, there’s only so much that a college can do to protect its students from faculty upheaval. The policy only covers scientific misconduct, not any of the other ways that a professor might depart — death, for example, or by taking a position at another university or in government or industry. In those cases, the students are still left to scramble.“It’s depressing,” says Goldman, “and perhaps it’s a flaw in the university system that’s developed over the last hundred and fifty years. But we’re doing our best to protect these students, because they’re the future of science.”

Competition at the Mall
The root of all science may be, as that plaque outside Bascom Hall suggests, the spirit of untrammeled inquiry, but the fundamental basis of research in today’s world is money — in particular, grant money.
This is what creates the temptation toward misconduct, and what makes protecting graduate students so difficult. Phil Anderson knows this well. In November 2007, he was going through the renewal application process for the one grant that keeps his lab running — a process that had consumed nearly all his time for more than two and a half months. The good news, he says, is that the application was finished; the bad news is that it was two weeks late, which, fortunately, the reviewers at the National Institutes of Health didn’t hold against him.A committee of perhaps thirty NIH scientists will judge that application, though it’s likely that only a couple will examine it closely. Though the committee members are competent researchers in their own right, the highly specialized nature of modern studies makes it unlikely that any of them will be as well versed in Anderson’s area of research as he is. That’s what creates the opportunity to falsify data on an application.“I’ve thought about Betsy a lot through this process,” he says. “What she did, I believe, happened because of the extreme pressure we’re all under to find funding.”

According to the National Science Foundation (NSF), UW-Madison is now the American university with the second-highest total of extramural research grant money — $905 million a year, when it’s all added up from every school, college, and department. That’s an average of $440,000 per faculty member, with some 90 percent of that money in the science and engineering fields. While that may seem like a lot of money, a great many things depend upon those grants, and the pressure to get them is severe. Take, for example, Anderson’s lab. The grant he’s applying for, an extension of the one that has funded his research for the last fourteen years, is worth $1.8 million during a four-year span. The university takes a share of this — some 48.5 percent — as overhead, what it calls facilities and administrative (F&A) costs. The rest, then, must pay for supplies: computers and microscopes and slides, not to mention C. elegans worms and the means to keep them alive and breeding. And it must pay for everyone who works in Anderson’s lab — his one assistant and three graduate students, covering their salaries and tuition reimbursements, and their own research, which will aid Anderson’s, but will also serve as the basis for the dissertations and other publications that will launch their academic careers.Should Anderson’s grant application fail, his salary will continue, but everything in the lab will shut down — the students and assistant will be laid off, the research halted, the worms discarded.“All of these things are paid for from outside sources,” he says. “In a way, the university is sort of like a big mall, and we’re all independent shops. We’ve got to come up with our own funding if we want to keep operating.”

But the competition is fierce and getting fiercer. There is a continual stream of new researchers — UW-Madison graduated 648 PhDs in 2005–06 alone — many of whom will seek funding for studies, and, as Anderson says, “you don’t get money for filling in the gaps; you have to be a pioneer” to be certain of winning a grant.In 2000, about one in every three applications for a grant from either NSF or NIH was successful. By 2007, that fraction had dropped to one in five. “For the most part, those are good scientists,” says Anderson. “They want to expand knowledge, and for most of them, there’s only one source of funding — the federal government. When they’re turned down, there’s no place else to go.”And that’s where the difficulty comes in for protecting graduate student whistle-blowers. It’s one thing for the university to guarantee that their funding won’t be eliminated should their professor be guilty of malpractice; it’s another to find the dollars to replace that funding once it’s lost. As he looks at pushing to make Irwin Goldman’s CALS plan a university-wide policy, Bill Mellon realizes that it will need more than morality. It will need money.“We’ll probably have to set up some sort of misconduct relief fund,” he says. “Fortunately, this sort of thing is very rare. But we’ll have to be prepared.”Such preparation is necessary, because the pressure to win extramural funding is so strong. Anderson believes that Goodwin is hardly the only scientist guilty of exaggerating to win a grant.“We’re all guilty of marketing ourselves, of casting our research in the light that makes it seem the most certain and revolutionary,” he says. “But the cardinal sin of science is misrepresenting one’s data. Openness, honesty, and fidelity to our data — those are what make science work.”

Forward
Two years after Elizabeth Goodwin’s resignation, her lab is still empty. It produces no income for the university, nor any research or educational benefits.Phil Anderson is still waiting to hear from the National Institutes of Health as to whether they’ll confirm his grant. His students’ futures — not to mention his own career advancement, and his research into nematode genetics — depend on the NIH’s answer. Irwin Goldman’s graduate student protection plan is likely to become university-wide policy this spring. “Morally, no one has any objections to it,” Bill Mellon says. “There may be a few tweaks here and there, but I expect it to pass this March.”At the same time, Mellon, who must shepherd the policy through approval by the faculty senate, has had his own grants to renew — his application for funding to study hormone action and vitamin D was due in February.And once Amy Hubert successfully defends her dissertation, she will walk across the stage and trade her nematodes for a sheepskin. According to Allen, Hubert is “a natural teacher,” and so she plans to take her career out of Madison and into someplace with less emphasis on lab results.“I’m hoping to get a position at a small college, where I can focus on teaching more than research,” Hubert says. “I’d still like to have my own lab, but I want to spend my time teaching and actually being in a lab, rather than holed up in an office writing grants. I don’t want that to be my whole job.”

John Allen is senior editor for On Wisconsin.

No comments: