Switching labs during a PhD

Jonathan Park is an MD-PhD student at Yale University in Newhaven, Connecticut.Credit: Jonathan Park


I could feel the tension setting in. I had just pressed Send on a flurry of e-mails to my advisers in which I’d included lines such as “I was hoping to meet ASAP to discuss potentially changing mentors and graduate program ….” And I couldn’t help but ask myself: was I making the right choice?

It was towards the end of 2017, and I had recently begun the PhD portion of my eight-year combined MD–PhD program at the Yale School of Medicine in New Haven, Connecticut. I couldn’t shake the feeling that changing laboratories could be a huge setback and risk, but felt I needed to switch so that I could gain the training I needed to achieve my long-term research goals.

I chose my initial lab during my first two years of medical school, which I started in 2015 after a bachelor’s degree in molecular biology. I wanted to learn bioinformatics and how to code, and I thought that having computational skills would be useful and empower me to work at the frontiers of personalized medicine. Joining a dry lab led by computational biologist Mark Gerstein was a logical move.

But during my clinical rotations, I took care of a patient with metastatic cancer who was relapsing despite his chemotherapy regimen. As I looked into alternative options, I became excited by the progress being made in using the immune system to treat cancer and developed an interest in immuno-oncology drugs, something in which Mark did not have research experience. I started to realize that our research interests were beginning to diverge.

I also wanted a wet-lab component to my PhD. I had worked in several such labs before medical school, and knew that some of my nascent research questions could be answered only through experiments. As I transitioned into my full-time graduate work in the middle of 2017, I came to the conclusion that another lab would better follow my career goals.

At first, I wondered whether I could fulfill these ambitions by staying in Mark’s lab. I really enjoyed working with him, finding him to be an encouraging and insightful mentor, and normally Yale’s PhD programs match an adviser to a student through their entire course. Contributing to a pioneering study of the genomic roots of neuropsychiatric disorders taught me a lot about the value of multi-institutional collaborations and integrative genomic analyses; moreover, I learned the core principles of how to do research from Mark. However, as we discussed my developing research interests, it became clear that I would not be able to pursue them in this lab alone.

He was supportive, and we discussed a potential co-advising situation. But my other academic advisers in the MD–PhD program warned against this, saying that unless there was a genuine collaboration between labs, it would create more administrative barriers on my path to graduation.

I then thought about which labs and mentors I might be able to link up with at Yale. I prioritized three things: first, a lab where I could use both wet-lab and computational skills; second, a research program in cancer immunology; and third, a principal investigator (PI) who had a track record of successful mentorship. This considerably narrowed down the list for me to three potential lab leaders.

I then gathered as much information as I could by scouring the faculty profiles on various Yale departmental websites and going through their publications. I sought input from my friends and peers. One PI, Sidi Chen, applied high-throughput CRISPR screens and systems biology to the study of cancer immunology — something that hugely aligned with my own interests. A friend of mine was already in Sidi’s lab, so I had long conversations with him about what the environment was like there. It sounded like a perfect fit for me.

At this point, I felt informed enough to reach out to Sidi and discuss the possibility of moving labs. My friend had already put in a good word for me, so during our initial meeting we mostly discussed my potential research projects and career goals. I felt encouraged. Sidi was mindful of my graduation timeline and interest in using both experimental and computational approaches. We had a shared scientific vision and belief in the future of gene editing and cancer immunology. He had space and funding for me. We seemed to be on the same wavelength. Everything clicked. I decided to make the jump.

The next couple weeks were a whirlwind as I formalized my switch after getting Mark’s blessing to move. He had always been supportive of my career, and leaving the place where I had grown so much for two years felt bittersweet. As I was also switching graduate programs from computational biology and bioinformatics (CBB) to genetics, I had to discuss coursework transfer with the directors of graduate studies. The fact that Mark was a co-director of the CBB program made this part easier. Forms had to be filled out, statements had to be written, and approvals from the deans of the graduate programs had to be obtained. There were many conversations with many different parties. Thankfully, everyone was working in my interest, and soon my switch was made official.

Of course, not everything started smoothly in the new lab. For the first couple of months, I struggled to gain traction on my projects because the learning curve was steep. But I found key people who helped bring me up to speed. Several postdocs served as ad-hoc mentors and collaborators, and trained me in various research techniques. I started to become comfortable again with running wet-lab experiments and coming up with projects independently. I fell in love with my thesis work and really believed in the translational value of our research direction, which was important to me as an aspiring physician-scientist. Looking back on the experience, now that I’ve graduated with a PhD, I can say that, at least for me, switching was the right choice.

Several months after joining Sidi’s lab, I met Mark to wrap up a couple of papers we were working on together and to let him know how I was doing. I always appreciated how graceful he was throughout the process. Changing PIs can be a setback, but I never considered the time I spent in Mark’s lab as lost. The computational training and ways of thinking I received under his mentorship would prove invaluable throughout my PhD.

Change can be stressful. The uncertainty can be palpable. Take care of yourself as you navigate the process! Graduate studies should be a time for creativity, exploration and discovery, and you deserve the best that it has to offer. For me, the risk is paying off and I find myself doing what I love. If you are at the same crossroads as I was, it just might be worth jumping into the unknown and charting a new course.

Portrait of Mark Gerstein

Mark Gerstein is a biomedical data scientist at Yale University in New Haven, Connecticut.Credit: Robert Lisak


When Jonathan told me that he was leaving my lab in late 2017, my initial reaction was one of disappointment. For most professors, having smart graduate students in their lab is a source of productivity and enjoyment. And having one leave in the early stages is, of course, a somewhat sad affair.

Jonathan had joined my lab at the beginning of 2016, with an interest in learning computational biology. He quickly became an integral lab member, working on projects that involved analyzing brain disorders and extracellular RNAs. During his two years in the lab, I could see his growth as a scientist and a computational biologist.

Given his progress, I initially tried to persuade Jonathan to stay. Sometimes there’s a small thing about the environment or personal interactions that can readily be fixed — such as changing the schedule of meetings, or taking a break from study. I also suggested a co-mentorship with another professor.

But after a while, I realized that Jonathan’s research interests had diverged from my focus on computational biology and were no longer aligned with those of the lab. He wanted to do something fundamentally different for his PhD.

At this point, my mindset completely changed: I decided to help, even encourage, Jonathan to move on to a different lab. This change reflects my experience over nearly two decades as a director of graduate studies for one of the PhD programs at Yale. If a student’s interests are not aligned with those of the lab they’re in, particularly in the early stages, they might do better in the long run to make a switch.

In contrast to the sprint-like coursework of undergraduates, graduate school is a long-distance marathon. An exam is often sufficient to motivate someone to study hard for a week, but this approach won’t work continuously for five years. The most challenging experience for many professors is getting an advanced graduate to see a paper through all the hurdles of publication and to properly finish a dissertation.

Advisers can’t push students to be interested in their projects; the motivation has to come from within. One of the best indications I’ve found of students being well suited to a lab at the early stages is that they read a lot about the field and come up with plenty of potential projects.

It is in the best interests of everyone that students are excited about their research. Over the years, I’ve certainly come across students who, I think, would have benefited from making a lab change at the beginning of their PhD journeys.

Often these individuals have initial enthusiasm. However, they show indecision about their precise direction and then become less interested in their chosen path over time. Paradoxically, this often affects those students who are stereotypically the best; these can often force themselves through the initial stages of a PhD, including the qualifying exam, even if they’re not that interested, but then lose motivation over the long term.

In this context, I think in particular of one student (let’s call him Chris) who rotated through many labs, had active discussions on which one to focus on, but was clearly undecided about his final choice until the eleventh hour.

He then stuck with his decision, breezing through qualifying, but eventually lost interest and drifted away from academia, going into the commercial world without completing his degree. He had no overt disagreement with his adviser or program — just a progressive loss of motivation that came, in my opinion, from a non-alignment of research interests.

The rationale for Chris to change labs was subtle, and distinct from the many more obvious reasons to change labs, which are easier to identify and address. These could include an adviser getting sick, leaving an institution or having a toxic personal relationship with a student.

In Jonathan’s case, it was clear that his research interests had shifted to cancer immunology, so I helped him move on to another lab. There’s one big qualification to my encouragement: we were lucky because of the structure of the program he was in. A conventional US PhD program in biological sciences presents no major administrative obstacles to changing labs in the early stages of a PhD, and doesn’t tie students’ funding directly to their lab.

This situation changes later in a PhD student’s career. Moreover, in certain other programs, a student’s funding is much more directly tied to a lab, or even to a project topic, from day one.

Given our luck with administrative issues, Jonathan’s transition went well. He moved from my lab on amicable terms and was able to finish two of the papers we had begun. Now, he has gone on to new topics and successfully finished his PhD in a different lab.

Leave a Comment