Dr. Gaël McGill
Founder & CEO, Digizyme Inc.
Faculty & Director of Molecular Visualization, Harvard Medical School
Co-author & Digital Director, E.O. Wilson’s Life on Earth
(bio @ Digizyme)
Sciencemotionology: Your background is very well-known, so it may be more interesting to know when you first started to imagine the events that take place inside the cell. Was it during your PhD preparation or even before?
Gaël: My interest in science and biology started at a pretty early age - it was in middle school that I became passionate about science. There were several factors that drove this passion, but I remember one book in particular that I kept reading over and over: Christian De Duve’s Guided Tour of the Living Cell. Similar to the Hollywood movie ‘Fantastic Voyage’ where one is shrunken down in order to experience the microscopic scale, De Duve asks the reader to shrink down to molecular scale and observe the cell from within. In fact each section of the book was organized according to a specific itinerary around the cell. For example, one needed to opsonize oneself to enter the cell through phagocytosis. If you wanted to see the inside of the nucleus, you had to wait for nuclear envelope breakdown during mitosis because otherwise you were subject to the nuclear pore’s gating mechanisms. It may sound obvious now in 2016, but this type of thinking really had an impact on me - I started thinking about molecules and cells in a very real, physical way. So I guess I would trace my interest in understanding how cells work back to that period.
Sciencemotionology: What are the major limitations to the way we picture the inner life of the cells? The almost total void of visual cues about the inside of cells? Or, perhaps the misleading indirect observations of cell organelles and proteins?
Gaël: This may sound trite, but I would say that the most limiting aspects have to do with spatial and temporal scale. Even with the pioneering work of those like David Goodsell to remind us about molecular crowding and the ‘geography’ of the mesoscale, or Drew Berry’s animations that remind us of the stochastic motion of molecules, we barely begin to comprehend the scales and speed with which any of this is happening. It’s great to have increasingly accurate measurements of these characteristics as research progresses, but the numbers still lie outside the scope of human intuition… 10 billion proteins in an ‘average’ animal cell? Each protein can cross the cell in about 1 second and encounters every other protein during this second-long random walk? Knowing it to be true does not necessarily enable me to ‘picture it’ in my mind! Hence the potential for visualization to give us a glimpse… to the extent that our visualizations (and the stylistic choices we make) reflect this information.
Sciencemotionology: When you started Digizyme, how was the landscape of scientific animation at that time? Who was your first client? Perhaps a colleague from your research group where you were doing your PhD?
Gaël: It’s hard for me to recall the overall landscape at the time because Digizyme felt like such a local experiment. We started the company in 1999 during my early graduate school years and I was lucky to have an incredibly supportive PhD mentor, David Fisher, MD-PhD. It was quite challenging in the beginning… I would do a full 10-12 hour day in lab and then come home to work on Digizyme projects. In those early days, Digizyme was not really a visualization company in the way that we think about it today: we focused on logos/corporate identity, websites, simple illustrations and vector/2D/Flash animations for the local academic, medical and biotech/pharma communities. I think our very first client was the amazing chemist Stuart Schreiber (Harvard University) in whose lab my wife Jeannie Park (software engineer, designer and Digizyme co-founder) worked in at the time. It’s only later - after designing and coding over 130 websites and animations! - that we migrated towards 3D visualization. We also began to realize the importance of developing better 3D software tools as well. The underlying premise or philosophy of the company, however, has always stayed the same: that there is great value in tightly integrating knowledge and passion for the scientific content with design and technology expertise. The result is an unusual team of scientifically-trained artists or artistically-trained scientists… whichever way you want to think about it!
Sciencemotionology: How hard was it for you to start doing animation? I think that at the time, a little bit like today, there was no formal training available for scientists who wanted to make animations. Is that so?
Gaël: That was entirely ‘trial by fire.’ In 2006, I suggested to my mentor Steve Harrison that Harvard Medical School (HMS) should have some sort of 3D visualization center. He agreed but funding for it was not available to create one. Instead he suggested that I start by teaching a new course to see if there was enough interest for this field at HMS. That conversation was in January 2006 and I had never touched 3D software at the time! I started teaching the course in April of that year… so I taught myself Maya during those 3 intense months. Luckily I was also able to recruit help from wonderfully talented people like Janet Iwasa (who was in Boston at the time) and my friend Eric Keller while I developed the curriculum for the class. Although it was quite a challenge to learn something as complex and deep as Maya in that amount of time, I relished the challenge of applying these advanced tools for scientific visualization challenges.
Sciencemotionology: Do you think you have an advantage in being associated with a university while doing animation?
Gaël: Not really. I basically have never done my visualization work as an HMS scientist. All the work I have done either personally or that my team has created has been through Digizyme. I try to keep the line between my academic and business life as clear as possible: at HMS I indulge my love of teaching and my federally-funded research focuses on how we learn from visualization, while at Digizyme I direct client-driven projects, create products in the science publishing space and design tools and resources for the community and education. It may be that some of my clients are aware of my ‘other life’ as an HMS faculty, but I don’t think it’s the reason that we get the projects we do.
Sciencemotionology: Client acquisition is perhaps the major challenge for animators and illustrators, or for any business, for that matter. Do you think you need to educate your clients about the advantages of using scientific visualizations?
Gaël: Over time I have found that it is not so much the difficulty of sourcing clients, but rather finding the right clients. By ‘right,’ I mean ones who are not too restricted budget-wise and who naturally appreciate that by working with us, they are not just paying for a product, but rather for the process they go through with us - I think that’s an important distinction. The most successful projects occur when the client approaches the design process with an appreciation for how the tough decisions at the intersection of aesthetics, storytelling, and scientific accuracy need to be balanced, and as a result they appreciate that Digizyme is their partner in finding this balance (not just a ‘vendor’). So although sometimes it is important to remind or educate clients as to the value of scientific visualization… I find the more important education is often in the realm of project management (i.e. things like when and how to provide the right kind of feedback). If you say ‘yes’ to enough of the ‘wrong’ clients, you can rapidly sink your business.
Sciencemotionology: What is your perception of how people see scientific animations? Do you think they feel it is something for kids? Do people, clients or not, realize the amount of work behind each animation?
Gaël: I think that very few people realize the amount of effort involved in scientific animations. People also generally underestimate how much scientific knowledge, research and preparation goes into a good scientific visualization. It’s not just the technical aspects of the 3D animation pipeline, but how much goes into truly understanding your topic and crafting the scientific narrative. Take for example the beautiful watercolors of David Goodsell - I think few realize that these are some of the more carefully researched visualizations around. I think of them as extensive ‘visual reviews’ of the literature.
Sciencemotionology: How do you feel about the idea of using scientific animations and visualizations in general to aid in research? To be part of the scientific method itself, to propose ideas, hypotheses about specific mechanisms - instead of being solely a communication vessel?
Gaël: This is one of the most important areas for progress to be made in our community at the moment: we need to explore this concept more deeply and identify case studies that showcase the intrinsic benefits of visualization to the broader research community. I think scientists are generally familiar with the explanatory (communication) and exploratory (dataviz and other tools) uses of visualization… but getting them to appreciate that the process in and of itself has the potential to shed new light on familiar data, provide common ground for interdisciplinary teams, seed more refined mental models and quickly reveal missing data remains a challenge.
Sciencemotionology: Do you still have the time to do animations yourself? What do you like most about the whole creative process?
Gaël: No… and I am very unhappy about this. In fact I have decided that the only way I can get back into this on a more regular basis, is through my teaching. Creating visualizations that exemplify a particular technical concept or showcase our tools (like our Molecular Maya kits) is probably going to be my ticket back into production. I miss it terribly and am generally unhappy when I stay away from the nuts and bolts of visualization for too long. As I noted earlier, I love the creativity that comes with identifying the perfect technical solution for a particular scientific visualization challenge. I also find that I can no longer read the scientific literature without constantly seeing opportunities for visualization. In fact it is my de facto ‘hunt’ and also my frustration… paging through the latest issues of Nature, Science and Cell every week is both wonderful and also painful in that I see all the potentially amazing visualization opportunities that we don’t have the bandwidth to create!
Sciencemotionology: How do you see the future of this field? Not only for the users but also for the creators? Do you think there will be a disruption in the way we make animations?
Gaël: It’s hard to tell. At the moment I am most focused on this idea of ‘democratization’ of visualization - for scientists but also for students. More intuitive tools can go a long way towards empowering those users who don’t typically create visualizations to now engage in the process and make their own images and animations. What if we could remove the ‘3D learning curve’ altogether? There will continue to be a need for specialized people at the forefront of the visualization field and I think that projects, collaborations and tools that call for that kind of expertise will continue to exist… but I would like to see how modeling a virus and rendering a beautiful image of it can change the understanding and motivation that a middle schooler can have for science. There is no better way for young learners to come into contact with data than to create visualizations of it.
Sciencemotionology: Recently you started Clarafi, a portal with resources and training designed for the makers of scientific visualization. What is your vision for this portal? What kind of reception has it had?
Gaël: Clarafi is a community of scientists, artists, teachers and students who share a passion for scientific visualization. At the intersection of the rapid pace of software development and the ever-changing landscape of scientific discovery, Clarafi offers training, tools and resources that help members improve their skills and tackle the unique design and technical challenges of science communication. In a sense, although Clarafi is a continuation of the site I created in 2006 - www.molecularmovies.com, my goal is to greatly expand the scope of training, resources and audiences. I also hope the site can become a platform to support the democratization of scientific visualization… getting new tools and skillsets into the hands of those who may not otherwise engage in this type of work. Eventually, the portal will be the home of several certificate programs that support not only visualization training for scientists but also visual literacy in the K-12 science classroom. So far the portal has been enthusiastically received even though it is in its infancy. Much more to come! Stay tuned :)
Sciencemotionology had the pleasure of meeting with Gaël McGill at the VIZBI conference that took place in Heidelberg last March and we exchanged some impressions about this amazing and colorful world of scientific visualization.
Founder & CEO of
"...to shrink down to molecular scale and observe the cell from within."
"...the numbers still lie outside the scope of human intuition..."
"...the tough decisions at the intersection of aesthetics, storytelling, and scientific accuracy need to be balanced..."
"...I think of them [watercolors of David Goodsell] as extensive 'visual reviews' of the literature."
"There is no better way for young learners to come into contact with data than to create visualizations of it."
Fantastic Voyage, 1966, by Richard Fleischer
Cellular Landscapes (detail) © Digizyme
Simian Virus 40 (detail) © Digizyme
Cellular Landscapes (detail) © Digizyme
Probes Ohmx © Digizyme