Legendary Labs: Secrets for Scientific Excellence

From management styles to creating the right culture, learn the secrets of academic researchers who produce impactful science and diligent scientists.

Published December 30, 2008

By Adrienne Burke
Academy Contributor

Phil Sharp, who won the 1993 Nobel Prize in Medicine and trained a scientist who won the same award 13 years later, says he learned from his first mentors how to nurture budding talent. While Sharp was still a grad student in chemistry at the University of Illinois, Victor Bloomfield gave his career a boost by telling other scientists about his work and by sending him to scientific meetings. And his postdoctoral advisor, National Medal of Science recipient Norman Davidson, encouraged Sharp to pursue his own research and engage with other faculty at Caltech.

As he continued his studies under 1962 Nobel Laureate James Watson at Cold Spring Harbor Laboratory, Sharp learned that “if you surround yourself with very exciting people and research projects in an environment where ideas are always percolating and you can add your own perspective, then it’s easy to do cutting-edge research.”

Sharp certainly makes it seem that way. Progeny of the MIT lab, where 30 years ago he discovered the split gene structure of higher organisms, now populate faculty posts at nearly every major university in the country. Sharp Lab alumni include Howard Hughes Medical Institute investigators, National Academy of Sciences members, and Andy Fire, who won a Nobel in 2006 at age 47. As a group, so-called Sharpies share such fond memories of their days under his tutelage that they organized 20- and 30-year reunions at the lab. Sharp counts them among the happiest days of his life.

Good Scientific Citizenship

Academic scientists such as Phil Sharp, who are as well known for producing excellent science as they are for developing following generations of top-flight scientists, are a unique breed. Within a system that gives recognition, money, and tenure for scientific achievement, good scientific citizenship generally goes unrewarded. Those who conscientiously nurture their successors’ careers are motivated by pure altruism.

And they are largely self-taught. Unlike in industry, where scientists in supervisory roles are typically immersed in management training, few universities offer even basic leadership instruction to newly minted principal investigators.

John Inglis is president of Cold Spring Harbor Laboratory Press, which publishes books on scientific management. He, says, “Postdocs who not so long ago did something really great and are given a lot of money and have to set about building a group are immediately faced with all kinds of challenges. Very seldom has anybody talked to them about how to do this leadership thing and how to cope with all the human situations that science throws up when you’re dealing with a creative endeavor.”

It’s no surprise then that the iniquitous university workplace—where senior investigators take credit for students’ work, schedule lab meetings on holidays, or provoke postdocs to hoard supplies and lock up their data by pitting them against one another—is no mere myth.

Carl Cohen, president of Scientific Management Associates in Boston and author of Lab Dynamics: Management Skills for Scientists, says lousy leadership is rampant in science. “Scientific projects get destroyed, interactions go astray, and students flounder, not because the science itself is wrong, but because scientists are not attuned to personal dynamics,” he says.

Searching for the Right Fit

Maryrose Franko, senior program officer for graduate science education at Howard Hughes Medical Institute, advises students against pursuing postdoc appointments based only on the principal investigator’s scientific accomplishments. Franko says many graduate students resolve to put up with whatever misery they must for the chance to work in a prestigious scientist’s lab. But, she warns, that strategy can backfire.

One promising young postdoc Franko knows signed on wittingly to the lab of a less-than-supportive P.I. “I warned her, ‘He’s a shark’,” says Franko. “But she said, ‘I don’t care, he’s the best in the field.'” Now, three years later, the senior investigator has prohibited the postdoc from taking her research to her first faculty appointment. She’s dependent entirely on a referral from him to get anywhere.

Kathy Barker, author of the popular lab management advice book, At the Helm, says that people frequently tell her that they wish their P.I. had taken a course or read a book about how to run a lab. “One in three people I talk to have had bad PhD experiences,” she says.

But does it matter? “The fact is that very great science can come out of groups that are disasters in terms of human relationships,” says Inglis. “A certain amount of money was spent, a certain number of people left science because they were so disillusioned about how the research enterprise works. But does any of that matter if the end result was a significant advance in our understanding of how a cancer cell works?”

Training First, Science Second

David Baltimore, past president of Caltech and Rockefeller University, would say it does matter. “I want to do great science, but that’s not the primary thing. The primary thing is the training, because that’s what’s going to last,” he says. To scientists like him, the advancement of the research ecosystem is more important than any single scientific discovery. And, as Barker points out, providing a future P.I. with an excellent experience can have far-ranging results: “Once you’ve been in a wonderful lab, you want to make your lab like that.”

Just what makes a lab wonderful? Even the most highly acclaimed leaders aren’t sure of the keys to establishing an excellent research culture. Says Sharp, “It’s sort of like cooking. You can follow a recipe, but you only know it works when it works.”

Asked to explain his secret to having trained nearly 100 accomplished scientists, including department chairs at Columbia, Duke, Harvard, and MIT, David Botstein says, “It’s a reasonable question, but I don’t know.” Botstein, who taught at MIT and Stanford before becoming director of the Lewis-Sigler Institute for Integrative Genomics at Princeton, says he sees it as his job to produce great students. But he has no formula. “I can only tell you what seems to work for me.”

Indeed, conversations with a dozen P.I.s widely recognized as great mentors reveal that few have any scripted approach to mentoring.

A More Guide than a Boss

George Church, director of the Center for Computational Genetics at Harvard Medical School, has launched some of the most promising young systems biologists in the country. Princeton Associate Professor Saeed Tavazoie, who zoomed from PhD thesis to tenure in just five years. So did Jay Shendure, an assistant professor of genome sciences at the University of Washington who was named to Technology Review magazine’s TR35 list in 2006 for a remarkable genome sequencing technology he developed in Church’s lab. But like most of his peers, Church candidly reports that he has never studied management or even thought much about it.

Nevertheless, these senior investigators have gleaned and put into practice a certain amount of lab management wisdom over the years. The advice they impart comes down to four simple maxims:

• Hire well;
• Be more guide than boss;
• Do your best to foster an open, congenial, collaborative culture; and,
• Put teaching and your underlings’ careers first, your research second.

While great lab leaders unanimously disdain micromanagement, hiring is one function they control carefully. “When you try to appear to run a laissez faire lab, you have few leverage points,” says Church. “The big one is whom you select. That affects tone, ambiance, and subject matters, so you need to exert quite a bit of certitude.”

Surprisingly, brilliance isn’t necessarily the first trait they seek in postdocs. “I don’t look for people who are very smart,” says Church. “If you got into grad school at Harvard or MIT, I don’t have to worry if you’re smart. I’m mainly looking for people who are nice.” Church says he is careful to not let his lab revolve around him, and he also shuns candidates who seem most concerned about their own success.

“Immersed” in Science

Phil Sharp looks for postdocs with a track record: “They’ve advanced a problem, can articulate what the problem is, and they have a view of the world that is developed and sometimes different,” he says of ideal hires. In grad students, he seeks those clearly “immersed” in science. “They read, they talk science, they work in the lab with a lot of commitment, and they go to lectures and come back with ideas.”

Bob Weinberg, a Whitehead Institute founding member and cancer research pioneer who has trained more than 100 scientists in his MIT lab, says his top criterion for selecting grad students and postdocs is that they be able to get along well with others. “I ask about that before I ask about scientific mettle,” he says. “How generous are they with their colleagues? How often do they share? I have turned down an applicant not because they weren’t brilliant, but because I’d heard they weren’t the most pleasant to have around.”

In fact, Weinberg makes it a point to survey candidates’ past mentors and labmates before making an offer. “You often have a postdoc around for three, four, five, six years. It’s kooky not to invest time in that detective work,” he says. “I don’t want people in my lab all to be in love, but I would like them to get along and share.”

Janet Thornton, director of the European Bioinformatics Institute in Cambridge, UK, goes a step further to have “a group of people who get on well together,” she says. She asks existing staff to screen incoming candidates, and takes their reviews seriously. When the feedback was, ‘if you recruit this person, the whole group will resign,’ she quickly rejected the applicant.

Pride and Ownership

“True grit” is what HHMI investigator Pippa Marrack looks for as “one of the best predictors of future success.” What’s true grit? “It’s about being brave enough to go for the core of the problem, and being persistent and not giving up when something looks like it’s going wrong,” says Marrack, who has run a lab at the National Jewish Medical & Research Center in Denver with her husband John Kappler for nearly 30 years. “It’s being able to persevere when the reviewers say your paper is crap.” Marrack says a one-day interview “can occasionally reveal when someone has done something in their lives that lets you see they have courage.”

Independence is another sought-after characteristic, especially among senior leaders with multiple responsibilities and busy calendars. “As my own life became more complicated running universities, increasingly over the years I have made independence a very important part of the equation,” says Baltimore. “The worst thing I can do is accept people who can’t handle independence.”

In fact, the freedom to pursue independent research is what most of these accomplished scientists say they most valued about their own training. “Dulbecco was a hands-off mentor, so I was given as much freedom to do what I wanted to and that made an impression on me,” says Bob Weinberg. Everybody in his lab has their own project, and knows up front that when they leave they can take it with them. That way, he says, “They can take pride and ownership in what they’re doing.”

Controlled Freedom

David Baltimore recalls that Richard Franklin at Rockefeller University “was a wonderful mentor because he gave me the freedom to do what I wanted to within the context of working on problems in virology.” Baltimore says the experience taught him the “tremendous importance of allowing young people to find their own way.” Over time, he says, “I’ve just developed great respect for what trainees can do if you support them and provide critical intelligence while letting them define as best they can where they want to go.”

Janet Rowley, the 1998 Lasker Award winner, famed for having identified a specific genetic translocation in leukemia, was mentored by 1966 Nobel Laureate Charles Huggins. She says that when she started up her lab at the University of Chicago in 1969 she approached lab management the same way she did child rearing. “You give people a lot of freedom, you’re there to help them if they need it, and you let them go,” she says. Rowley also says she prefers flexibility to rules and regulations: “You don’t know where creativity is going to come from, and as lab director you have to be open to it coming from an unusual direction.”

George Church’s lab is so free of rules that he compares it to an artists’ colony. “I couldn’t be in a cookie cutter mold where the lab was real production-oriented like a factory, or so hung up on dogma and protocol that you couldn’t think outside of the box,” he says. Having been trained in a research environment that rewarded creativity and interdisciplinary effort, Church says he has adopted the same system.

Equality for People and Ideas

Church is also a fan of equality for people and ideas. “I try to treat everybody as a peer. The lab isn’t entirely without hierarchy, but it’s historically been pretty flat.” As a result, it operates like a free-market system. “If I want to get something done, I have to sell my idea down the line. If it doesn’t sell, I realize there’s something wrong with my message or it’s a bad idea,” Church says.

While all of the scientists interviewed for this article talk about the importance of being supportive to their trainees, they also all see value in letting people flounder and learn from their own mistakes. Church says there’s a fine line between maintaining a nurturing environment and one that promotes critical thinking. “You don’t want to be so supportive that you can’t tell someone something is a bad idea, but you don’t want to be so critical that they think all their ideas are flawed or that all good ideas come from one person.”

Joan Steitz, James Watson’s first female graduate student at Harvard, has run a molecular biology lab at Yale since 1970 and been an HHMI investigator since 1986. She says surprising things can happen when a P.I. steps back and lets postdocs work on problems they’ve developed independently.

Shobha Vasudevan, a PhD who joined Steitz’s lab from the University of Medicine & Dentistry of New Jersey, came along with what Steitz thought was a rather dull research project on protein binding. But when Vasudevan started finding proteins associated with microRNAs and showed that microRNAs can activate genes depending on the cell cycle, Steitz says the project “went off in the most amazing direction.” Vasudevan’s paper, which Steitz calls “very important,” was scheduled to appear in Science before the end of 2007.

Creating a Congenial Culture

Across the board, successful leaders abhor the idea of promoting competition among scientists inside the lab as a way to stimulate discovery, saying collaborative, open environments are the most productive. “Most of the truly original ideas that have come from people in my lab come from frequent conversations with their peers,” says Weinberg. “I want them to be talking incessantly. I want that to be part of their style—to be talking to people outside of my lab.”

In that vein, Pippa Marrack says her training at the renowned MRC labs in Cambridge taught her the simple importance of eating in the cafeteria to learn about others’ work. “Everybody, all the Nobel Laureates and the janitors, ate in the same cafeteria at large tables there,” she says. Marrack set up her students’ offices to promote conversation. “We keep our postdocs and grad students in two large offices without cubicle walls to encourage them to talk to each other and come up with ideas together. They do sit around talking, and it’s not always about fantasy football,” she jokes.

Weinberg says he has spent years trying to make sure the limits of his students’ universe are not the walls of his lab. “We’ve been having floor meetings since 1970 with six or seven groups where we all share our research findings. I want people in my group to talk openly about their successes and failures so they can benefit from others’ insights. They may have to go much further afield than me to get input. I want them to develop this habit rather than leading hermit-like existences.”

The Road to Experimental Research

Phil Sharp says he chose MIT as the home for his research because it offered that opportunity for interaction with other scientists. “I’ve been at MIT 33 years, and on the fifth floor of the cancer center all that time,” he says. The laboratories of Weinberg as well as David Housman, Michael Yaffe, and David Sabatini are all nearby. “We live in a group,” Sharp says. “We share a noon Wednesday seminar, we have a party every Friday afternoon, and we have science talks together. My lab is immersed in a group of about 100 people, and in that group are some of the best people in the country in every age group. It elevates us all.”

Another reason a congenial culture is important: the road of experimental research is a rocky one. “More often than not, things don’t work,” Weinberg says. “How do you maintain morale when things aren’t working? I’m not saying I’m the personification of morale maintenance,” he says, “but I think I’ve created an environment where people can help each other through the scientific rough times, if not the personal ones.”

While networking is seen as a benefit, thoughtful mentors tend also to keep their own labs small enough that they can contribute to the success of each person in it. In the late 1980s and early ’90s when funding was at its peak, Rowley says she had as many as 16 lab members. “That’s really as much or more than I could carefully manage. When a lab is 40 or 50 postdocs, I question whether a senior investigator can really counsel that large a number of postdocs carefully, creatively, effectively. I look on very, very large labs with a certain amount of skepticism,” she says.

Produce People First, Science Second

David Botstein says that he has always considered his profession as a geneticist to be not simply research, but a hybrid of teaching and research. “My goal with students and postdocs was in part, of course, to do research that would be of general interest, but also to choose problems and methods that would maximize the students’ learning,” he says.

Botstein argues that putting teaching first is a key not just to generating better scientists, but to producing better science. “The time I spend teaching—up to half my time—makes my research better,” he says.

Rowley agrees. Supporting a young scientist’s success reflects well on a senior scientist, she says. “If you keep asking yourself, ‘What am I getting out of this?’ you reduce your effectiveness. You have to really think about what is going to help the other person be more successful.”

Phil Sharp’s method for helping students develop their careers is to get them to take ownership of an idea and then to plan and execute a series of experiments that advance the understanding of science in that area. “It works best if that interest aligns with my interest in the lab. Occasionally it will align to something only distantly related, but I’ve always found it most important to put the person’s development at top.”

Let Them Taste Success

Church steers postdocs in directions that are most likely to let them taste success. “It’s a soft touch,” he says. “It’s amazing how little it takes to steer, but you don’t want any of your postdocs doing something that is so impossible that there won’t be milestones or they won’t get any credit.”

Even before coaching his postdocs on the problems they chose to tackle, Bob Weinberg sees an important role for himself in influencing their thinking. “I want to impart to them a taste for working on problems that are important and will be thought to forge new conceptual paradigms.”

In discussions over lunch at least twice a week, and in a journal club where they analyze recent scientific publications, Weinberg teaches his people to think critically about research questions. “Is this an interesting question?” he challenges. “Have they focused on something important, or is it trivial in terms of its heuristic value? Are the data really that interesting? Or are they just filling holes in a brick wall?” He says these questions train people to think about whether or not a topic is worth the investment of time.

The Hardest Thing in Science

Baltimore considers “framing the right question” to be “the hardest thing in science.” He says questions have to be audacious enough to be interesting and yet experimentally tractable. “Finding that balance of interest and do-ability is something you only develop with experience and with trying things that are too hard or doing things that are not interesting enough,” he says. “I try to help people find that sweet spot. And when we’re successful, they do wonderful things and they develop a lot of self-confidence, and when they leave my lab they’re ready to establish their own labs and be successful.”

Contrary to the supervisor of the hapless postdoc who has no rights to the data she produced, Weinberg makes it a point to ensure that the research his postdocs do in his lab will help launch their careers. He says, “Some labs have rules that when they train grad students or postdocs, the project stays in the lab. When my lab continues in an area of research, I try to stay out of the way of the person who has gone away so they’re not being undermined by my lab.” To do otherwise, he says, would be very unfortunate. “I’m interested in their soaring, not sinking,” he says.

Also read: Grant Rejection Could Be the Best Thing for Your Career and 10 Things To Do at Every Scientific Conference