A BIOLOGIST TALKS to a PHYSICIST Working Alongside People from Different Fields Can Help in Improving Communication Skills and Identifying Gaps in Knowledge

Advice, technology and tools

Send your careers story Your Work to: naturecareerseditor story @nature.com THINGLASS/GETTY A BIOLOGIST TALKS TO A PHYSICIST Working alongside people from different fields can help in improving communication skills and identifying gaps in knowledge. By Kenneth S. Kosik

s a physician–scientist, many of my exclamation mark. River, practically an ocean away from the colleagues were surprised when I When I made the move ten years ago, I had physical-science labs at Harvard University in moved my laboratory from the Boston wanted to shift my biomedical work closer to Cambridge, Massachusetts, on the other side. Longwood campus at Brigham and the interests of physical-science researchers, So, engaging scientists in physics and chemis- Women’s Hospital in Massachusetts to discover broadly applicable principles try, or in computer science and engineering, Ato the University of California, Santa Barbara within the framework of biology, and to grasp was challenging — especially for a medical (UCSB), where there is neither a medical school the multilayered complexity hidden in nearly doctor trained rather narrowly in molecular nor a university-affiliated hospital. every question posed by biologists. and cellular biology. More than a few e-mails arrived — some The physical sciences are not lacking in My goal was simply to open a conversation expressed puzzlement, some surprise, some Boston. In fact, they are world class. But my and possibly a collaboration with physicists, had a wink, but all were punctuated with an lab was on the ‘medical’ side of the Charles not to become one. As a relatively small

Work / Careers institution with a distinguished faculty in both the physical sciences and engineering, UCSB was, I felt, an ideal place to wade into this territory. Over the ensuing decade, this risky move has resulted in my sharing many graduate students and postdoctoral fellows with computer-science, engineering, physics and chemistry faculty members. These collab- orations have greatly broadened the science in all my publications. I devised a few simple rules to help the biologist in me to cross the divide between the life and physical sciences. In learning to talk to physicists, I discovered that I can com- municate better with everyone and clarify for myself what I do and do not understand in my own field.

Understand what ‘I do not understand’ means. When physicists say they do not understand something that you have said about biology, it’s possible that you do not understand that topic either. ‘Understanding’ operates at different planes in different disciplines, and when a physicist seeks understanding, what they hope to

grasp might differ from the knowledge that FERNANDEZ SONIA a biologist seeks. Neuroscientist Kenneth S. Kosik. For example, a biologist understands gene transcription by identifying specific transcrip- Deep computational knowledge is a huge tion factors, their binding sites, the role of RNA asset for biologists, but many have not had polymerase and the genes that get activated. the opportunity to hone these skills or even be For the physicist, these crucial facets of The fastest way to introduced to them. In my experience, open- transcription — specific gene names and make a physicist’s eyes ing a conversation and sharing students with binding sites — are extraneous details. Instead, physical scientists is step one towards bridging among the questions they consider important glaze over is to recite this gap. are the probability distributions associated biological jargon.” with attracting transcription complexes to Don’t be flummoxed by physicists’ maths. specific sites and the quantification of the In discussing their own work, physicists will forces involved in this process. occur according to a clock? Identifying these often reach for a formula. After they write Clearly, questions of this nature are of keen unknowns in biology is more stimulating than the equation and stare at it as if pondering a interest to biologists once we sidestep our love recounting textbook-level information. Mark Rothko painting, they might proffer an affair with our favourite gene. explanation. You can have a productive con- Recognize the posture of false modesty. I have versation without the bread-and-butter basics Seek common ground. When a physicist says found that physicists often display a false mod- of physics, such as Hamiltonians or modelling they do not understand an aspect of biology, esty regarding their knowledge of biology. A an Ising phase transition. Bear in mind that they are not requesting a ‘biology 101’ expla- physicist will say biology is much more difficult physicists can be more sensitive about the nation. In my experience, when physicists ask than physics because there is a lot to memorize. maths than are mathematicians, who are not a biology question, they want to apply the They might say: you know an encyclopaedia of bound by reality and see the world as a con- thinking of physics to biology; specifically, detail, and all I know are a few equations. straint. they are searching for universal, mathematical Albert Einstein said: “As far as the laws of explanations. Keep in mind the maths shortfall in biology. mathematics refer to reality, they are not Physicists move away from settled ques- For most biological phenomena, we don’t have certain; and as far as they are certain, they do tions. In biology, much less seems settled. precise equations — unlike in physics. This is not refer to reality.” The biologist must have Emphasizing what you know is less interesting not to say that we don’t have maths, but our substantial comfort with uncertainty. Com- than saying what you need to learn. field needs a lot more detailed quantification. pared with the good predictions from theory Many of the missing pieces in biology are This lack is the Achilles heel of biology, and in physics, prediction in vast facets of biology, quantitative details, such as the absolute copy affects even the concepts we use every day. such as evolution, is more like stock-market numbers of a protein or an RNA in a single cell It is a rare biologist who can explain the cycles — discernible only in retrospect. that kinetically mediates some function, and negative binomial used in RNA sequencing, how the cell keeps track of so many regulatory parameterize organelle shape and distance Scale matters. Consider this matter of dials. Do cells perform regular maintenance scales when assessing active transport and dif- perspective: most of the time, what a physi- on a parts-replacement calendar, as is done fusion, or interpret the phase diagrams that cist does is much smaller or much bigger, or for aeroplanes, or is damage the only trigger correspond to biomolecular condensates — a much colder or much hotter, than anything for replacing parts? Does damage tend to surging interest in biology. we biologists do.

Biologists should know the sizes of can frame a problem in these terms, it will element. If you are uncomfortable engaging macromolecules — the organelles and the probably engage a physicist. with specialists in fields in which you are not cells that they work with — and first consider trained, or if you don’t want to delve into topics how they scale (or do not scale) with systems Physicists laugh a lot. Not only is the humour that might not immediately bear a clear con- derived from physics. of physicists arcane, but almost anything nection to your established domain, then the unexpected can provide a jocular moment. course I took is not for you. However, if you are Consider precision. Physicists require much Theirs are the ultimate inside jokes, which are not only comfortable with this approach, but greater certitude to draw a conclusion. They often not obviously funny. But laugh along require stimulation from outside your imme- might be a bit smug when you claim signifi- anyway — even if you don’t find the humour, diate realm of investigation, there is nothing cance with a P value of less than 0.05. For them, they won’t know the difference. better than to look in unexpected places. attaining a result with a P value of 0.05 is like the chance of hitting a barn door with a base- The future Kenneth S. Kosik is a professor of ball: you can’t miss it. To conclude, let me say this path is not for neuroscience at the University of California, everyone. Knowing yourself is the crucial Santa Barbara. Avoid jargon. The fastest way to make a physicist’s eyes glaze over is to recite biological jargon, such as gene lists and pathways. Come to think of it, it’s the fastest way to make any- one’s eyes glaze over. A PHYSICIST TALKS Skip some details. Physicists do not want to know about all your controls. They want the concepts, and are willing to assume you have TO A BIOLOGIST performed the appropriate control experiments and have chosen the correct methods. If The fastest way to understand a new discipline is you’re doing a ‘chalk talk’ (illustrated presenta- tion) for physicists, be sure to emphasize your to embed yourself in it. By Sarah Bohndiek ideas more than detailed data and techniques. Physicists make judgements on the basis of clear, cogent, compelling ideas. t’s amazing how many of biologist Ken As I progressed through my studies, I realized Kosik’s points in his accompanying careers that I wanted to gain a deeper understanding Manage expectations. Physicists expect you column (page 281) resonate with my expe- of cancer itself and how we might exploit its to build your own equipment. If this topic rience as a physicist working in the oppo- biology so we could use non-invasive imaging comes up with a theorist, feel free to ignore it. site direction. I enjoyed my training in to find the disease. If it comes up with an experimentalist, you will Iphysics largely because it didn’t involve rote I hadn’t studied biology since I was 16, so have to concede that you paid a huge mark-up learning (needed by biologists to get that I started looking for postdoctoral positions for off-the-shelf equipment. encyclopaedic knowledge) and it allowed me in ‘friendly’ environments — biophysics, bio to derive much of the material from a few key engineering and so on. But I soon realized that, Understand optimization versus the ‘good- equations and principles. to truly understand cancer, I needed to break enough’ principle. Biological processes carry In his column, Ken notes that when a phys- out of my comfort zone and immerse myself evolutionary baggage, and therefore arrive icist says they do not understand something in that environment. at solutions that might not be optimal. If, in biology, they are not requesting a lengthy My first postdoc was in Kevin Brindle’s for example, you were designing a machine ‘biology 101’ explanation. But I do sometimes biochemistry laboratory at the University of that walks, it might not be best to model it want a biology 101 refresher so I can gather Cambridge, UK. I learnt how to hold a pipette, on human walking, which is not necessarily a deeper understanding of the biological run a magnetic resonance imaging system, the most efficient way to get somewhere. In problem while I think about how to address it. conduct animal experiments and design biol- the absence of evolution, the physical world Ken also talks about being uncomfortable ogy studies. is often assumed to operate according to with uncertainty. Personally, I can be quite In return, I fixed temperamental magnets, optimized parameters. By contrast, biology comfortable with uncertainty when I can take wrote hardware-control code and super- operates in what has been called a Wadding- steps to control it; it’s just that, in biological vised some of the lab’s more physical-scienc- ton terrain: processes function well even when systems, I normally have to accept that I can’t. es-oriented projects. After three years, I had they are not as efficient as possible — as long a much greater understanding of cancer bio- as they are still good enough to work. A deeper understanding chemistry and how to conduct in vivo imaging The PhD I received from University College studies, but I was missing exposure to clinical Consider a physicist’s perspective on reduc- London in 2008 was in radiation physics, with applications and connection to my previous tionism. Reductionist approaches to complex a focus on evaluating the potential of CMOS research home in optics. biological systems do not necessarily mean image sensors (the silicon behind most smart- My second postdoc, with Sam Gambhir in reducing those systems to small units. Systems phone cameras) for application in medical the department of radiology at Stanford Uni- approaches to biology that involve large num- X-ray-diffraction studies. versity, California, took me back to my roots. I bers of variables can be inherently interest- Most of this research took place in a dark explored new projects in optics, trying to make ing to physicists because they conceptually room, where I examined the optical response imaging with optics faster and cheaper than resemble problems in statistical mechanics of the image sensors, but because I was in a before and opening new horizons in fields such that relate microscopic and macroscopic prop- department of medical physics and bioengi- as endoscopy. erties with very large degrees of freedom — and neering, I was exposed to a range of biomedical As part of Stanford’s Molecular Imaging Pro- exact solutions are not possible. If a biologist challenges — including cancer detection. gram, I saw teams build new technologies in the

Work / Careers lab and take them through to first-in-human by teaching physics in undergraduate or studies. I was inspired by Sam’s dedication graduate courses. to revolutionizing early cancer detection and his incredible passion for transforming Do not blindly accept dogma. Challenging technology into clinical trials. That passion prevailing ideas in biology using your per- still drives me in my current role at the Uni- spective can bring about revolutions. I greatly versity of Cambridge, UK, where the mission admire colleagues who have upturned dec- of my interdisciplinary team is to advance our ades of accepted dogma using quantitative understanding of tumour evolution using methods that were not even considered by the next-generation-imaging sciences. We oper- biological community. ate between the department of physics and the Cancer Research UK Cambridge Institute, Perfect your pitch. Working across disciplines and this gives us access to a community of means you’ll inevitably be talking at confer- optical physicists as well as cancer biologists, ences and meetings to groups of scientific and a translational pipeline through to clinical non-specialists; the best communicators application. perfect a balance of generalist and specialist material in presentations and make their Build synergy language accessible to all. I’m a technology geek at heart. I love to dive into the underlying physics when I design optics Avoid equations during presentations. It’s and to explore the engineering of new devices; not just biologists who would prefer not to but, for me, the story is never complete without see them. Illustrate equations with graphs and seeing these methods through to application. visuals as much as possible. If you do have to I feel that the research path I explore today is a include equations, write them on a board in

synergy of all my experiences. real time so your audience can follow your BOHNDIEK SARAH I am often asked what I learnt as I traversed Physicist Sarah Bohndiek. guided explanation — don’t just flash them this path from physics to medicine. The fol- up on the screen. lowing are the key steps I had to take during my own personal development, which pre- Find a good mentor. Finally, and most impor- pared me well to embrace my interdisciplinary tantly, I would not be where I am today without career. A strong element of my incredible mentors. When you move disci- academic achievement plines, you need people who can help you to Learn the language. Just as with a foreign integrate into a new community, show you how language, the fastest way to become fluent in is community to navigate a new funding landscape and advise the language of a new scientific discipline is to recognition and on expectations for junior faculty members, embed yourself inside it and practise regularly. networking.” when the emphasis on publication and impact For me, learning the language was the most varies dramatically between disciplines. important reason why I moved from physics to biochemistry for my first postdoc. Ask questions. I had to, and there are no stu- The future pid ones. I also attended introductory lectures More unified training, encompassing physics Get comfortable being uncomfortable. I went for first-year undergraduates to fill in gaps and engineering applied to biology and medi- from being one of the most knowledgeable in my knowledge. Going back to school and cine, will help to equip the next generation of people in my research field at the end of my admitting what you do not know is crucial. researchers with the skills they need to operate PhD to knowing less than most of the first-year across these disciplinary boundaries. PhD students in my new lab. For me, that meant Embrace uncertainty. I continue to be amazed I still advocate for specialist training in one I was doing something right, but you do have to by the complexity of biology and hence the field as an undergraduate and then broadening be OK with taking that hit and spending time uncertainty in results. Unfortunately, it’s likely research horizons gradually through advanced building a new skill set. that I’ll never be as certain about anything in academic training. biology as I can be about new discoveries in A strong element of academic achievement Don’t forget that you bring unique skills. physics. is community recognition and networking; I had a lot to learn along my journey, but I moving too rapidly between fields can make brought useful skills at each stage. My quan- Learn statistics. Having a quantitative back- it challenging to build a reputation for excel- titative mindset was valuable in designing ground doesn’t necessarily mean you really lence. Imaging is kind in this respect: regard- imaging experiments, my engineering skills understand statistics. I had to dive back less of how you form an image, there are many were useful for fixing broken equipment and into statistics to design the best biological commonalities in the research that mean it is my bottom-up physics approach to problems experiments. relatively easy to tell a story within your career offered a fresh perspective. and maintain connections along that journey. Don’t lose touch with your roots. Being able Building that story and your profile will Test biological-research hypotheses. Some- to continually innovate at the interface of ultimately be crucial, no matter what path times you’ll bring a quantitative perspective physics and biology means staying grounded you choose to take. that enables you to test a hypothesis or embark in the physical sciences, but this can be hard to on a research topic that wasn’t accessible do when you are in a biological department. To Sarah Bohndiek leads an interdisciplinary before. Don’t be afraid to ask the biological compete for faculty positions, you need to be team combining physicists, engineers, question yourself rather than waiting for your able to teach your subject, so find ways to keep chemists and molecular biologists at the collaborators to provide an answer. it fresh in your mind, as I did, for example, University of Cambridge, UK.