Dr. Todd Golub is a world-class innovator in integrating genomic methods into cancer research and medicine. This background has enabled him to lead The Broad’s set of exhaustively interdisciplinary labs, which research diseases including schizophrenia, tuberculosis, and many others. In addition to directing The Broad’s cancer program and serving as Chief Scientific Officer, he has made trailblazing discoveries in the molecular basis of childhood leukemia. Dr. Golub also researches cancer at the Dana-Farber Cancer Institute and teaches pediatrics at Harvard Medical School. Dr. Golub received his B.A. from Carleton College and his M.D. from the University of Chicago Pritzker School of Medicine. In our interview, Dr. Golub shares with us benefits and pitfalls accompanying automation in research.

Business Today: Could you tell me about your experiences from your early years as a student and researcher that led you to your current career and work at the Broad?

Dr. Todd Golub: I started doing research as an undergraduate at Carleton College, and you either get bitten by the research bug or you don’t. To me, the details of methods didn’t matter so much, but the idea of discovering something that had never been discovered before was exciting. So, I spent a lot of my free time while in college, during summers between years, and during breaks doing biomedical research.

I then decided to go to a research-oriented medical school for the same reason – because research-oriented students would be supported and encouraged, and there would be lots of role models for future physician scientists. After, I went to do research-oriented clinical training in pediatric oncology. But it wasn’t until my postdoc after all my clinical training that I had a long, fully protected block of full time research effort, where I focused on using what was at the time modern genetics and genomics to study the genetic basis of one of my patients’ leukemia. There was an unusual chromosome translocation in the leukemia cells, so we hypothesized that if we could discover the genes that were broken by that chromosome translocation, that would provide biological insight into why that kid’s leukemia developed.

Today, that sequencing experiment lasts a couple weeks, but in the early nineties that was not straightforward at all – it was a several-year undertaking. There was no human genome project at that time, so it turned out that one of the genes that was broken by this translocation was a gene that hadn’t yet been discovered. But discovering a gene in the human genome doesn’t happen anymore, because we have the complete human genome sequence. So that brought me to a faculty position at Harvard Medical School and the Dana Farber Institute, and that then led to my collaboration with the Whitehead Center for Genome Research, which was the forerunner to the Broad Institute.

BT: I was reading that you had a formative experience working at a small lab that granted you intellectual freedom. With the Broad being such a large and very different institution, how has this small lab experience impacted your approach to research today?

Golub: I don’t think the size of an organization matters. What matters is how wide open you encourage the lens to be amongst people in your lab. My view is that the lens should be very wide, and that people shouldn’t feel constrained too much by either prior knowledge or discoveries, or by their position or status in the research ecosystem. People should feel like they have the ability, the freedom, and the expectation to do something really important.

That happened for me in a small lab in Chicago – where at the time I didn’t realize it was small. I was excited about discovering something really cool and important, and the freedom that I have to think big started there. And those principles don’t change when you get to a different lab.

BT: You mentioned that rank doesn’t matter, and that attitude encourages a very inquisitive atmosphere. How else do you foster this inquisitive culture of research? How do you push scientists to ask creative and different questions? Or is it all up to scientists’ own respective interests?

Golub: I think a lot of creativity comes from within. But I think one thing that we can do as mentors and leaders is to encourage people not to be limited by what they see as certain constraints. By that I mean too often scientists, especially young scientists, start a project by thinking about how much money they have available to do their project, or when they need to get the project finished, or what have other people done already, or what do we even know about the particular problem.

This is a different mindset than thinking about what is needed to crack open the field; what would be important. After that, we deal with issues like whether we have the technical feasibility to pull it off or do we have enough funding to pull it off. Those should be secondary considerations, which might indeed lead to not being able to do the project. But if you put those secondary issues first, you’re dead in the water.

BT: Molecular Biology research is becoming increasingly automated, with high-throughput screening experiments and assays capable of testing hundreds of chemicals in a matter of minutes. What nascent automation technologies in cancer research do you think are the most important, promising, and fascinating?

Golub: Certainly, high-throughput DNA sequencing is having a big impact on being able to characterize human tumors. That used to be a very low-throughput process, and now it can be highly automated in a factory-style sequencing operation. CRISPR technology and other genome-editing methods can be scaled up to look at the entire genome at once, by for example mapping out all genes in the genome in one experiment. That is very powerful.

I would emphasize that I don’t think these breakthroughs are about automation per se; I think they are more about the underlying molecular biology that makes it possible to think big, and to do experiments at scale. And once it becomes interesting to do that, then it is worth investing in the automation to scale up according to peoples’ imagination. But it is not that the automation itself has created some new capability. In fact, much of the automation that is used at the Broad is not all that sophisticated. So, if you focus only on automation capabilities, you wouldn’t get very far.

In fact, for many projects, trying to automate the entire process results in slowing down the research rather than accelerating it, because it may be that if there is a step that can be done easily by a human you should just do it by a human. And so, slavishly trying to fully automate an entire process is often a misdirected effort.

BT: The Broad has pioneered a new model of scientific research that incorporates scientists across disciplines. What else sets the Broad apart from other research institutions? What is the Broad’s next step in creating innovative research methods?

Golub: I would say that many institutions have multidisciplinary research, so I don’t think the Broad can lay claim to that. I think the Broad has become very good at bringing scientists coming together to take on something together that they couldn’t accomplish on their own. In some cases, that is multidisciplinarity. Sometimes it’s not being afraid of research of a certain scale that would be more difficult in academic labs. Sometimes it’s taking on a project that just doesn’t make sense for a company to do because the research should be a public good, and the monetary value cannot be extracted by a company and therefore it makes no sense for a company to do.

So, I think we work in this very rich ecosystem of biomedical research, and I think the Broad has specialized in trying to identify and execute on those projects that will help crack open a field, but don’t make sense for a company to do because it is best done as a public good. And yet, our research can’t be tackled by an academic lab because it requires a certain infrastructure, scale, management, and integration across multiple disciplines, in particular laboratory-based data generation and computational analysis. Having both of those disciplines together is very powerful; having them separate can work in principle, but in practice that doesn’t work as well as with them together.

So, I think the unique problems we tackle are what distinguishes us from other labs.

BT: What was your proudest moment in your career?

Golub: I can’t point to a single proudest moment. When I see people that have trained in my lab flourishing as independent investigators and continuing the culture of the lab that puts a premium on making a difference in the world over personal career achievement, that makes me very proud.

BT: What were the unique strategies developed in the Broad’s founding?

Golub: A very substantial part of the Broad’s founding was empowering young people to think big and empowering them to lead. This is not always customary in the corporate world or the academic world.

When I started leading what became the cancer program at the Broad Institute, I had just started my lab as an assistant professor. And so, I was an unlikely suspect to lead anything new or large. That was part of the Broad’s culture from the beginning: identify good people and give them encouragement to think big, and it’s okay to give them authority to lead. And so, in an effort among four principle faculty – Eric Lander, Stuart Schreiber, David Altshuler, and myself – working with the universities Harvard, MIT, and the Harvard hospitals, we created a structure that would become the Broad institute.

The main purpose of creating that structure was to help ensure that other scientists who wanted to come together spontaneously could do so. It is an umbrella administrative framework to allow grassroots activities to flourish without administrative boundaries getting in the way of scientific progress.

BT: What advice do you have for undergraduates pursuing research, medicine, or other science fields? How do you prepare for constant change in this industry?

Golub: My advice would be to focus on what is going to have the biggest impact. If that requires developing new technologies, then establish the collaboration needed to develop those new technologies. If making the biggest impact involves throwing out the technology you just finished developing, you need to be prepared to do that as well. If you keep your eye on the ball of maximizing your impact in a way that is going to move the needle, I think you’ll be just fine.

I think people get lost chasing specific papers to be published, or trying to follow what they think is hot in the field rather than stepping back and not skating where the puck is but skating where the puck is going. You need to have enough flexibility of mind so you don’t paint yourself into a scientific corner. If you can do that – and it’s not easy – then I think everything will be fine. ﹥