By turning once-overlooked health information into data sets, Michael Waskom ’05 is helping researchers create targeted treatments for cancer
As a senior data scientist at the pioneering health technology company Flatiron Health, Michael Waskom ’05 works at the leading edge of cancer research. He builds machine-learning tools that draw on real-world data to expand treatments and improve patient care. Waskom comes to oncology from a background in computational cognitive neuroscience. Before joining Flatiron, he was a research scientist and Simons Fellow at New York University’s Center for Neural Science, where his work focused on the neural basis of higher-level cognitive processes, such as learning and decision-making. He earned his bachelor’s degree at Amherst College and his doctorate at Stanford, where in his spare time as a first-year graduate student he created Seaborn, a popular data-visualization library for the programming language Python. We caught up with Waskom via Zoom from his home in New York City, where he’d just returned from a run in Central Park with his border collie, Darwin.
How are you using data science in the fight against cancer?
There’s lots of information that is captured in the routine course of care of cancer patients. In the past, that information might have been scrawled on a piece of paper and put in a filing cabinet. Nothing about an individual patient’s experience led to any changes in how cancer is treated. Flatiron takes the raw, unstructured information from oncology clinics and academic medical centers across the country that is now captured in the electronic health records system—doctors’ notes, lab results, test results—and we turn it into data sets that can be analyzed to improve treatment and develop therapeutics.
Your academic background is in computational cognitive neuroscience. What made you decide to move from research science to health tech?
Academic research is incredibly important and exciting. I don’t have any doubt that the kinds of things I was working on are relevant for understanding, for example, the neural basis of disorders like depression or autism. I have a lot of confidence in that. But one thing that can be frustrating is that outside of a few narrow domains, such as building brain-machine interfaces for quadriplegics or other people of limited mobility, it can feel like we’re a long way from amelioration. Maybe within the next few decades we’ll have a really good model of what’s going on in the neural networks of people with autism, and then we’ll be able to create technology for shaping brain network activity that will improve people’s lives. Cancer is much farther along in the process.
Where is cancer research headed?
It’s an exciting time. The other day I was looking at a plot of approval of new targeted therapeutics that are designed to attack a particular protein expressed by a particular tumor in some patients. Advances like these are leading to significant improvements, step changes in median survival times. In some cases the targeted treatments are also much more tolerable than standard chemotherapy. It’s not just taking a huge hammer and whacking your system, and hoping it kills the cancer cells first. The slice of the population that a particular cancer drug is going to treat is increasingly targeted. Three percent of patients have this mutation and it can be targeted with this drug, a different three percent of patients have this other mutation. So, step by step, the data tools we’re building can find each of those three percent and determine what is the right drug for each.
It’s fascinating how many disciplines converge in your work—neuroscience, molecular biology, computer science, oncology, clinical practice. How do you prepare for a career like that?
I think where we’ve been successful in using technology to improve things in the world is not just from having people who are incredibly focused on one esoteric topic. You need that, but there also have to be people who can take those esoteric insights and connect them in the world. I think a liberal arts mindset helps to produce that second group of people. One thing I remember about Williston is that my senior year I did a spring project on the mathematics of music with Greg Tuleja. It was an opportunity to pursue a line of inquiry that was more independent than a formal course, and I think it has echoed through other things I have done, an interest in seeing connections between disparate fields. At Amherst I started out as a philosophy major and became interested in philosophy of mind questions, which pushed me toward psychology and neuroscience. I ended up creating an interdisciplinary major, and then after graduating I went to work in a neuroscience lab at MIT. I’m glad I studied philosophy, because it trains you to think rigorously about abstract concepts, which is so much of math and machine learning and data science.
What do you do in your free time?
Right now I’m training Darwin to recognize shapes. I picked up one of those little wooden shape sets for kids, and he’s building up the idea that each of the blocks has a name. I say a name, he goes to it, he gets a treat. He’s pretty good about triangle and oval. He’s still learning pentagon. It’s fun. Also, it’s instrumental in keeping his brain occupied so he doesn’t eat the furniture.