%20(1).jpg)
For Pat Zhang, it was a certain curmudgeonly but brilliant fictional television doctor that sparked his interest in healthcare and led to him pursuing pre-med courses as an undergraduate. As a 12-year-old, Wes Field enrolled in a summer course at Sacramento State University — “Intro to Virology” – spurred by what he calls a “morbid fascination with the Ebola virus.” But, in fateful turns, both happened into computer science courses and were surprised to find their calling. “I was immediately drawn to the problem-solving nature of working with data,” Zhang recalls. “And I could impact lives by taking the healthcare route within data science.” Today, Zhang is a data scientist and Field is a principal machine learning engineer on Amgen’s Digital Health & Innovation (DH&I) team—whose goal is to make data-driven discoveries that may lead to medical breakthroughs—supporting Amgen in delivering on its mission to serve patients.
One of the world’s leading independent biotechnology companies, Amgen has supported millions of patients with serious illnesses through innovative medicines focused on six therapeutic areas: cardiovascular disease, oncology, bone health, neuroscience, nephrology, and inflammation. Data scientists at Amgen work in cross-functional teams that can include biologists, clinical experts, physicians, engineers and commercial leads. “To arrive at a solution, we work with a cross-section of stakeholders to define the problem, acquire and engineer the data for modelling, and share the model with the larger team,” says Field. “Ultimately, we’ll deploy successful models to benefit Amgen’s patients.” Within the DH&I team, “there’s a great diversity of thought because of our wide-ranging backgrounds and perspectives,” adds Zhang.
Zhang and Field find their work to be gratifying, as well as relevant to our particular moment in time: last year, DH&I was charged with formulating an algorithm using a de-identified data set of U.S. Covid-19 patients that could predict risk scores based on medical histories. “The one-week hackathon included a cross-functional team using different techniques,” says Zhang. “With a data set of 80+ million patients, there was a great deal of potential in what we could learn.” The effort included four sub-teams using different modeling approaches; Field led the group that used binary classifiers to predict Covid mortality. “The impressive thing is that each of our four sub-teams identified roughly the same comorbidities,” he says. “When we shared our results with the CDC, they confirmed these comorbidities were being observed in real time. Since this was early in the pandemic, every bit of consensus was valuable in treating Covid patients.”
Both appreciate having opportunities to get involved in projects that reflect their interests. Zhang recently worked on Amgen’s Crystal Bone project for patients with osteoporosis and co-authored the accompanying publication. The team used deep learning models based on historical electronic medical records to help doctors predict the likelihood of fracture over the next two years; this way, they can begin preventive measures to preserve bone health. “Currently we’re working to bring this tool to healthcare networks to see if our results generalize to a broader setting,” he says. Field’s most meaningful project is the work he is doing with Amgen’s Clinical Development team, focused on Next Generation Trial Execution. “The goal of the project is to optimize how Amgen selects clinical trial sites based on trial criteria and historical data,” he says. “Running successful clinical trials is paramount to serving our patients—for example, what may seem like small gains in the speed of our trials means getting more new medications to the people who need them, sooner.”
Amgen has long championed data science and its role in developing healthcare therapies and tools. “The ethos here is that data science and machine learning can improve any part of Amgen’s business,” Field explains. “I don’t know of any other company where you might be working on antibody optimization one month and syringe defect detection the next, but Amgen makes this possible with rotations for new data scientists.” Zhang points out that Amgen also supplies ample resources for continued learning in the field. “Before the pandemic, I spent three months at deCODE Genetics, a world-renowned research center in Iceland where they have genotypic data for all of the country’s citizens.” deCODE uses its population-wide data set to identify variations in the human genome associated with common diseases and develop targeted therapies. “I was there with other data scientists from Amgen to apply artificial intelligence techniques to use the data in different ways,” he adds.
Zhang joined Amgen straight out of graduate school, not entirely sure what to expect from the biotech setting. “I was pleasantly surprised by the fact that while Amgen is not a tech company, our DH&I group sometimes feels like it is,” says Zhang. He notes that Amgen’s headquarters in Thousand Oaks, California, has “a start-up vibe, with colorful furniture, ping-pong tables on some floors, and free coffee.” While work has been remote over the last year, teams are still in close communication through tools such as Slack, where they can retain some of their in-person culture. “We’ve created one channel for the sole purpose of expressing gratitude; you just shout out someone if they do something that helped you and you want the wider team to know,” says Zhang, beaming. “There’s also a healthy use of memes.” Field started at Amgen during the pandemic: “I was nervous about integrating into a team that was so close and collegial, but those concerns were immediately put to rest by how warm and inclusive the teams at Amgen were,” he shares. “And they’re some of the smartest people I ever met.”
Looking at the potential of data science in the future, Zhang hopes that it can be used to address inequities in healthcare. “Part of that will be to ensure that clinical trials have—at a bare minimum—diversity that matches the overall population, and then ensuring these patients have access to subsequent care,” he says. Field is also excited about the possibilities of data science and healthcare. “Every project at Amgen has real weight and impact,” he says. “Our data science culture is shaped by that shared value—ultimately, we are people who want to positively impact our world.”
Are you a data scientist looking to make a difference? Visit careers.amgen.com to learn more about working at Amgen.
This story was produced by WIRED Brand Lab for Amgen
