On the Case at Mount Sinai, It’s Dr. Data



Jeffrey Hammerbacher is a number cruncher — a Harvard math major who went from a job as a Wall Street quant to a key role at Facebook to a founder of a successful data start-up.

But five years ago, he was given a diagnosis of bipolar disorder, a crisis that fueled in him a fierce curiosity in medicine — about how the body and brain work and why they sometimes fail. The more he read and talked to experts, the more he became convinced that medicine needed people like him: skilled practitioners of data science who could guide scientific discovery and decision-making.

Now Mr. Hammerbacher, 32, is on the faculty of the Icahn School of Medicine at Mount Sinai, despite the fact that he has no academic training in medicine or biology. He is there because the school has begun an ambitious, well-funded initiative to apply data science to medicine.

A tall man with deep-set eyes and a close-cropped beard, Mr. Hammerbacher stood recently in front of a white board filled with an amalgam of computer and genetic code, speaking of “Linux clusters” and “gnarly C code” — standard terms in the language of computing — but the main subject was biomedical science. Mr. Hammerbacher also discussed “neoantigens” and “gene variants,” and the data-driven hunt to find and understand the rogue cell clusters of cancer.

“We’re pursuing problems that are computationally and intellectually exciting, and where there is the potential to change how doctors treat patients in two or three years,” Mr. Hammerbacher said.

Eric Schadt, the computational biologist who recruited Mr. Hammerbacher to Mount Sinai, says the goal is to transform medicine into an information science, where data and computing are marshaled to deliver breakthroughs in the treatment of cancer, Alzheimer’s, diabetes and other chronic diseases. Mount Sinai is only one of several major medical schools turning to data science as a big part of the future of medicine and health care.

They are reaching out to people like Mr. Hammerbacher, whose career arc traces the evolution of data science as it has spread across the economy. After a job designing trading models at Bear Stearns, he worked for a few important years at Facebook, where he started the social network’s data team and made his reputation and a tidy sum. Next, he was one of four founders of Cloudera, a fast-growing company that makes software tools for data science. And now he is immersed in medicine.

Getting to this point has not been a predictable journey. He grew up in the Midwest. His father, Glenn, was an autoworker at a General Motors plant in Fort Wayne, Ind., and his mother, Lenore, was a nurse. He has had to overcome challenges, including mental illness. He has stepped beyond his background — the data of his life — more than most people.

Jeffrey Hammerbacher was always a numbers guy. That’s evident in a school paper he wrote as a 7-year-old in Fort Wayne. “My favorite hobby is doing math while I’m eating,” he wrote in clear block letters. “I like doing this because math is my favorite subject and I like to eat.”

Math is still the subject he most admires. “It’s snobbery on my part, but I view math as the true arena in which human intellect is demonstrated at the highest level,” he said.

He wasn’t only a mathematician during his childhood. He began reading at age 4 and was soon one of the Fort Wayne library’s best customers. He aced the test that allowed him to attend an excellent private school on a scholarship. Rachana Fischer, who is now a litigator in Silicon Valley, was a year ahead of him in high school. She recalls a recalcitrant student, but one who had read not only every math book in the school, but also the works of poets like Frank O’Hara and Vladimir Mayakovsky.

“His career is based on analyzing people by data and numbers,” she said, “but he doesn’t fit into any box himself.”

Mr. Hammerbacher was a star pitcher in high school and, while his grades lagged, his athletic ability and his scores on standardized tests got him into Harvard. But by his sophomore year, he was struggling. He rarely attended classes, and he simply skipped final exams in his sophomore year.

Looking back, Mr. Hammerbacher called his behavior “very wasteful” and “dumb” and part of a disturbing pattern. “I’ve had troubles with mood and anxiety at times in my life,” he said. “That’s a challenge for me. It’s been very humbling.”

He left Harvard and returned to his family’s home in Fort Wayne, where he joined his father for a while at the G.M. factory. A year later, chastened by his experience, he went back to Harvard, and graduated in 2005.

From that point on, Mr. Hammerbacher says, his career has been a matter of repeatedly following the smartest people in search of the best problems. Many of the sharpest math graduates from Harvard were going to Wall Street, and he did too.

He landed at Bear Stearns in the summer of 2005 — three years before it collapsed in the financial crisis. As a quantitative analyst, he built sophisticated computer models, mainly for mortgage securities. “On a single mortgage hedge,” he recalled, “we made or lost more in a day than my father made in a lifetime.”

He stayed at Bear Stearns less than a year, enjoying the intellectual challenge of the work, but ultimately finding it unsatisfying. “Our whole goal was to make the models more complex,” he said. “It was a really bad use of quantitative skills.”

Better problems, Mr. Hammerbacher believed, were on the horizon. He wanted to work for one of the Internet companies that were becoming natural laboratories for data science, and in 2006, he joined one of the most promising. It was a Silicon Valley start-up with fewer than 50 employees, including several people he knew from Harvard. He got a job at Facebook.

The social network was then a two-year-old company with sky’s-the-limit ambitions. Mr. Hammerbacher was 23, two years older than Mark Zuckerberg, Facebook’s founder and chief executive.

Soon after arriving, Mr. Hammerbacher decided he wanted to use Facebook’s data to improve its service. With a small team, he built software tools for gathering, analyzing and experimenting with data. Mostly, these were tests of what works best. For example, which page layout or feature change prompts users to spend more time on Facebook, or makes them more likely to send messages or post pictures.

One group of online users is shown the change being tested, and another group is not. This so-called A/B testing is routine now in website development, online advertising and marketing. At Facebook, it would become more sophisticated in the years after Mr. Hammerbacher left. But he played an instrumental role in starting it. “He laid the foundation for doing data analysis at scale,” said Itamar Rosenn, manager of data science infrastructure at Facebook.

Facebook was a great petri dish for data science, but after three years, Mr. Hammerbacher tired of social networks and online advertising. He departed in late 2008, with shares worth millions, but with far less money than he would have made if he had stayed until Facebook went public in 2012. He has no regrets, he insists. Leaving Facebook, he said, freed him to “work on things that matter more.”

The first thing he did was become a co-founder of Cloudera, which was valued last year at $4.1 billion. He retains the title of chief scientist, although he now spends most of his time with Mount Sinai.

While at Cloudera, Mr. Hammerbacher says, he took the punishing start-up lifestyle to self-destructive extremes.

Marathon workdays were capped by nights of partying, drinking and recreational drugs. His professional and personal life suffered. At work, Mr. Hammerbacher’s appearance at meetings became less predictable.

In March 2010, his old mood and anxiety problems erupted into a full-blown crisis. Mr. Hammerbacher was rushed to the emergency ward of a San Francisco hospital with a raging panic attack that felt like cardiac arrest or a stroke. A psychiatrist later diagnosed bipolar disorder and generalized anxiety. He quit recreational drugs and alcohol. He was treated with prescription drugs for several months, and also found that regular sleep, eating habits and exposure to sunlight helped.

Today, by all accounts, Mr. Hammerbacher is healthy and fit.

“He really committed himself to being sober,” said Halle Tecco, his wife, who is executive managing director and co-founder of Rock Health, which funds and advises start-ups in health care, “and to his family, his friends and his work.”

Mr. Hammerbacher responded to his personal health crisis with a characteristic outburst of curiosity, reading and seeking people to learn from. That propelled him toward medicine — and to work with Dr. Schadt at Mount Sinai.

Dr. Schadt had concluded that medicine was ripe for a data-driven revolution. Chronic diseases, Dr. Schadt explained, are not caused by single genes, but are “complex networked disorders” involving genetics, but also patient characteristics such as weight, age, gender, vital signs, tobacco use, toxic exposure and exercise routines — all of which can be captured as data and modeled.

“We are trying to move medicine in the direction of climatology and physics; disciplines that are far more advanced and mature quantitatively,” he said.

That message resonated with Mr. Hammerbacher. By 2013 he was spending most of his time in New York rather than on the West Coast, assembling a research team that now numbers 10 people. Their expertise spans the breadth of data science: machine learning, data visualization, statistics and programming.

His group’s objective is to alter how doctors treat patients someday. For example, Mount Sinai medical researchers have done promising work on personalized cancer treatments. It involves the genetic sequencing of a patient’s healthy cells and cancer tumor. Once the misbehaving gene cluster is identified and analyzed, it is targeted with tailored therapies, drugs or vaccines that stimulate the body’s defenses.

Mr. Hammerbacher’s team does not do the basic science. Other researchers do that. His group works on the “computational pipeline,” he said, with the goal of making personalized cancer treatments more automated and thus more affordable and practical. “It’s ultimately what cancer cures are going to look like,” he said.

The road to technology revolutions is paved with failure, halting progress and hard work. Mr. Hammerbacher and his colleagues are engaged in pathbreaking yet often frustrating data science work.

He is optimistic about his initiative’s prospects, but has come to appreciate that the mysteries of the human body may be more resistant to math than finance or social networks are. Today he speaks less about quants taking over than about their lending a hand. “We’re not the most important people,” he said, “but we can help.”


Check Also

Mental Health Tips and Resources For Doctors

May is Mental Health Awareness Month and as a Doctor, it’s important to take care …