By: Deepti Mathew
In a period of merely five years, the field of data science has grown from a niche for specialists to a red-hot job option for freshly minted biomedical PhDs with just a touch of tech savvy. While the six figure salaries have undeniable appeal, the opportunity to partake more directly in the implementation angle of science is also a feature that appeals to many a jaded bench scientist. Since staring at data and spotting trends are what many scientists-in-training do anyway, why not complement it with the much-coveted basic understanding of coding, in order to land what has famously been billed “the sexiest job of the 21st century”!?
If you are asking the question “WHAT is a data scientist?”, you would have to have been living under a rock! Yet the HOW, WHY, WHERE, and WHEN of transitioning from Scientist to Data Scientist eludes many of us. This piece covers just that. To provide an insider’s picture of what it looks like to be one on a day-to-day basis, four real-life data scientists sat down with aspiring PhD students and Postdocs at the “What Can You Be With a PhD?” career symposium (WCUB). This event, convened biennially and organized by a consortium of 15 New York City research institutions, has become a popular venue for nascent scientists exploring non-traditional career paths, both within and outside academia.
The Many Paths to a Data Science Career (a.k.a. HOW?)
The introductions of the panelists’ career storylines made it clear right off the bat that a PhD student can take different routes to destination “data science.” Anasuya Das trained under the popular Insight Data Science Fellowship, to jump start her career at Memorial Sloan-Kettering following an unsatisfying stint as a neuroscience postdoc at NYU. Tom LaGatta, a charismatic mathematics PhD, learned all about machine learning and big data on the job at Splunk, a company that specializes in data visualization and analytics for pharmaceutical firms, among other clients. Priya Kar took the MBA route to transition from an academic career spanning cancer biology and leading research projects, to being a healthcare professional on the business side of things at Bristol-Myers Squibb and Signals Analytics. The panel facilitator, Matthew Oberhardt, a data scientist who chose to work within a hospital setting at New York Presbyterian Hospital’s Value Institute, shared his stepwise foray into computational approaches to biology, spanning right back to his PhD days. All except the youngest panelist opined that their postdoctoral training was positively instrumental, though not absolutely essential to landing their current job, reflecting the newer trend of PhDs eschewing the traditional postdoctoral training and looking for ways to directly work towards their careers of choice.
You Can Do This, Yes You! (a.k.a. WHY?)
One of the most defining moments of the session came when the panel requested members of the audience who were skilled at data science to raise their hands. Barely any did! The panelists then expressed surprise that a hall filled with certified scientists dealing with data everyday did not consider themselves well-versed in both “data” and “science.” The point was that you probably already have what it takes to be a data scientist. And what is more, you’re well-positioned to learn the technical skills you don’t yet possess. The question is, are you driven enough to seek avenues to supplement the skills we all already possess? “Be brave” was one WCUB panelist’s heartfelt advice. The business culture can be cultivated and statistical know-how can be picked up, but irreplaceable is the resilience and analytical mindset one develops during their scientific training. “Aligning the business question with the science perspective is really hard,” explained Dr Kar, “and that’s where a PhD comes in!”
Now that we’ve established why we qualify for these positions, let’s address why we should want to consider this career path. Quality of life was discussed as one of the primary motivators. The added bonus was that while the pay was generally high, one can decide just how high one wishes to rise based on how highly one prioritizes work-life balance alongside monetary gain. One panelist shared anecdotal evidence of an NYC data science colleague who lived on a farm in New Jersey complete with horses in the stable, whilst another used work-from-home privileges to write code from a tiny café in Manhattan’s trendy East Village.
Another motivating factor that seemed particularly valuable to the panelists was the fulfillment that arose from their sense of being more closely involved with the scientific endgame. All panelists reported that their current jobs felt more aligned with the real-world application of science than that of their former laboratories.
The Data Science Umbrella (a.k.a. WHERE?)
Data Science, much like cancer or the flu, is sometimes an umbrella term common to many different types of jobs, all of which are multidisciplinary, and one is better equipped to find their niche as they progress. The attendees learned that while they would always be the subject-matter specialist in their firm (“the oncology expert/the biomedical expert”), where exactly they are going to bring that subject matter to life, using their analytical or data science prowess, can vary considerably. Dr. LaGatta was full of enthusiasm while describing his discovery of the benefits of the Sales Engineer role, while Dr. Das outlined the kind of career trajectory she was hopeful of achieving after starting off in an Analyst position. Dr. Kar championed big pharma as a great fit for data scientists with a flair for strategizing, while Dr. Oberhardt’s role leaned towards optimal management of hospital data.
The end product depends on where under this umbrella one fits. A day in the life of a hospital data scientist, for instance, will most resemble an “internal consulting” job and entails creating dashboards, interfacing with clinicians, helping them to understand data and thereby make hospital decisions. The panelists agreed on consistencies in these well-oiled corporations, regardless of the product. Different corporate functions may hold appeal to different individuals, and by developing an awareness of the workflow in their workplace, the budding data scientist can eventually decide what floats their boat. In fact, finding the corporate function that you need to plug into to feel you make the maximum impact, was touted as the single most important factor while determining where your data scientist career could take you.
It was further emphasized that data scientists are sought after in a wide range of industries including scientific organizations, medical and research institutes, tech companies and consulting firms; budding data scientists can find or reaffirm their broad area of interest once they get a foothold in the field. A glance at the nature of the employers represented by the four panelists cements this notion.
Getting Started (a.k.a. WHEN?)
As goes the popular adage, the best time to start preparing for your transition to data science was at the start of your PhD; the second best time is now! The panelists discussed some of the early steps they took and offered advice on how a scientist-in-training can best use their limited time to effect their transitions out of academia. While knowledge of complex computer programming languages isn’t required, familiarity with basic coding is highly desirable, and is, moreover, not as hard to come by as you might think in the age of free online courses. Coursera was mentioned as an example of a useful resource. Taking advantage of career development clubs universities offer and talking to people outside one’s lab were highly recommended to begin exploring avenues that your career might lead you to. Sometimes it’s simply a matter of probability. As Dr. Das put it, “When you try more things, you’re more likely to discover stuff that you find a bit of fun and are reasonably good at too.”
Every type of personality offers opportunities to chart one’s journey. As a postdoctoral fellow, the measured and methodical Dr. Oberhardt made a chart with tables listing the features of the careers paths he was considering, which he then updated based on informational interviews. This helps elucidate the kind of things you care about as important criteria. Moreover, the clarity this brings shows in one’s responses when they interview. Dr. Kar, on the other hand, favored following one’s gut instinct and making leaps of faith that are right for you, as she did when deciding to pursue her MBA and to opt for strategy and operations at Bristol-Myers Squibb.
At the stage of interview prep, familiarizing oneself with industry jargon is one effort where a little can go a long way. While the lingo is often just vague terminology that shouldn’t overwhelm you, it does indicate to interviewers that you’ve done your homework regarding the field of data science and machine learning. To this end, the panelists recommended reading up to stay abreast of what’s trending in medical research and familiarizing oneself with the landscape of the pharmaceutical industry. One could learn, for example, what the companies are working on and who their rivals are (for BMS, Merck!). Willingness to learn was deemed more important than subject matter expertise; let this shine through in your interview. Upon landing the job and settling in, the advice for the next stage was to find the corporate function that makes sense to you and find how that function of your role drives the mission of that corporation. Get comfortable with thinking of yourself within the corporate context and the data scientist in you will be best positioned to thrive!
In summary, data science is indeed a happening field. Hopefully, this coverage of the WCUB data science panel brought to you by the NYC SciComm team, can help illuminate how you can make it “happen” for you!
About the Author: The author is a New York based PhD scholar who recently leveraged her LinkedIn network to transition from cancer research to a career in promotional medical education. When not tending to patient samples and feeding cultured cancer cells, she can be found experimenting with pressure-baking to tend and feed her 4-year old, who just so happens to be a seasoned world traveler. Connect with Dr. Mathew at linkedin.com/in/deeptimathew .