About

Technology and science; petabytes and petri dishes; algorithms and alleles. This is a story of the blending of seemingly distinct but surprisingly harmonious disciplines. It unfolds within and around the walls of Novartis Institutes of BioMedical Research (NIBR) in Cambridge, Massachusetts. But this total rethink of the practice of medicine is happening across all of Novartis’ labs and research facilities throughout the world, and has the potential to impact every one of us.

Scroll to start

Dr. Wendy Winckler was looking for something. Something really small. A molecular glitch inside a cell.

Before her glowed the data collected on tumor samples from patients with colon cancer. One of those samples revealed an unexpected genetic mutation, although something about it was familiar. Dr. Winckler immediately recognized it as a driver of some breast cancer tumors; more importantly, breast cancer tumors that were responding to their prescribed treatment.

It was 2006 when Winckler made this observation. She was working as a research scientist at the Broad Institute of MIT and Harvard.

At that time, oncologists were unlikely to have tested their colon cancer patients for this genetic sour note.

Genetic sequencing was still new, slow and pricey.

This unexpected connection between the colon cancer tumor and breast cancer tumors fueled Winckler’s decision to dive deeper into cancer genetics. She now leads the Next Generation Diagnostics group at Novartis Institutes for Biomedical Research (NIBR) in Cambridge, Massachusetts. Her team is helping reframe how cancer is studied, understood and treated.

What Were You Doing in 2006?

Interactive tour created with images from Google Street View and set to motion using Teehan+Lax Lab's Hyperlapse tool.

Fast-forward to a recent pilot test of a successful breast cancer drug. Winckler, who was not involved in the study but is intimately familiar with it, explains that the drug was being tested with bladder cancer patients. “The overall outcome…was negative,” she says, sounding a little more excited than one might expect. Overall, the drug didn’t lead to improved survival for the participants. The trial failed.

F
F
F
F
F
B
F
F
D
F
C
F
F
B
F
F
C
F
F
D
F
F
D
F
F
F
B
F
F
F
B
B
F
F
F
F
D
F
F
F
C
B
F
F
D
F
F
B
F
F
F
C
F
D
C
F
F
F
B
F
B
F
F
F
F
F
C
C
F
F
D
F
F
F
B
F
C
C
F
F
F
F
F
A+

One patient, however, had an “exceptional response.” The tumor disappeared.

It’s obvious how that single, lucky patient might consider this “failure” a success. But to grasp its impact on a much larger scale, you need to understand how technology is changing the process—and outcomes—of research and discovery.

“Every genome is telling a story. There’s something buried in the data that we’re looking at.”


Cost of Genome Sequencing in

"It's just not obvious to us. Finding that answer, that piece that makes sense of it, is what keeps us going." Preethi Srinivasan teases genomic stories out of Big Data at NIBR. She’s a Bioinformatic Scientist in Dr. Winckler’s Next Generation Diagnostics group.

Both women point out that Big Data has radically changed genetic research. Up until a few years ago, high costs limited the amount of sequencing research teams could run: one or two genes for a typical study, ten for a big one. This meant teams would have to make their best guess about which gene might be involved in the cancer. A lot was riding on that one sequence.

Today, your genome can be sequenced for a song.

New technology is hitting the market this year that could bring the cost down to a mere $1,000 a pop. That’s better than 99% off the original cost—bargain basement pricing! Winckler explains, “Now that we can generate these tremendous data sets...we can be unbiased. We can say, ‘Let’s test all the genes to see which one it is.’ It’s very satisfying to know that we are getting away from the same old hypotheses that we’ve been looking at for 20–30 years.”

Remember that colon tumor Dr. Winckler was studying at the Broad Institute? If its owner were diagnosed today, just seven years later, he or she could feasibly undergo comprehensive gene testing. And the results could suggest a different treatment; one that could incorporate breast cancer therapies.

It sounds massively impersonal: Big. Data. But data sets big and small allow technology to create a more personalized experience of life for us based on what we do—how we shop (Target, Amazon), what we listen to (Pandora, Spotify), what we share (Facebook, Foursquare). It also saves our butts on a regular basis. How else would we land jobs (LinkedIn), plan travel (Orbitz), find true love (eHarmony)?

And now, Big Data is allowing science to create a more personalized experience of health for us based on who we are—at the most basic, molecular level.

Our unique mutations send each of us marching to the beat of our own drum. They’re also pointing to a new, highly personalized approach to the treatment of what ails us.

As it turns out, we are our data. And our data may save us.

Take the bladder cancer patient who benefited from the breast cancer drug. The genetic data of that patient’s tumor revealed a mutation that’s involved in the molecular pathway that the breast cancer drug targets. If researchers could find this mutation in other patients’ tumors, it’s possible those patients also would respond to the drug, no matter where their tumors originated.

Wait. What?

One highly targeted drug. Combating multiple diseases.

No wonder Dr. Winckler was so excited by what would otherwise be considered a failure. “This Big Data study on a single individual opened up a whole new area for patients who have the same type of genetic events.” That single patient’s data could save others. It could save you.

So exactly how much data are we talking about? Stored in a computer one patient’s gene sequence takes up 5-8 gigabytes. That’s about the size of a typical iTunes library, which is manageable. But for discovery, scientists are looking across hundreds of people and multiple diseases.

“These data sets aren’t even human-readable files. You need to have a team with diverse expertise—computer scientists, biologists, clinicians—working in close collaboration on every problem,” explains Winckler. “We’re going to find tens, hundreds of mutations in every single person. Some of them are contributing to the cancer but many of them are not.

“Trying to sort out the signal from the noise is much more challenging in these large data sets.”

Jim Bullard Algorithm Slinger On the agony and the ecstasy of wrangling Big Data (1:04)
Preethi Srinivasan Data Whisperer On the DIY Big Data career path (1:56)

The more data you have, the more connections you can make, right? All you really need is some whiz-bang data scientists and software engineers to wrangle all those letters, lay some beautiful algorithms on them, and wait for the magic to happen.

But wait, that’s not how science works. Scientists wear white lab coats and work with petri dishes and test tubes. True enough. But biopharmaceutical companies like Novartis see the importance of assembling collaborative, multidisciplinary teams that bring an array of expertise and insight to the table, not to mention a variety of coats.

At Novartis, discovery teams are composed of biologists, software engineers, computational biologists, laboratory scientists and statisticians. They’re all searching for genetic variations that are related to illnesses, working with fragments of genomic data, maybe 100 base pairs long.

Each fragment is like a musical bar. The teams try to figure out where each bar fits in the larger symphony of the human genome.

Know someone in college or the tech industry who’s looking for the most promising field in which to sling their 1s and 0s? Two words: data science. It’s the skill set that’s all the rage in scientific circles. And it comes with the added benefit of potentially saving people’s lives.

Data scientists come from all walks of very smart life. Some are strong in math and physics, while others excel at creating algorithms. Some bring experience from the physical sciences, others from software engineering. Achieving that diversity in their teams is critical at Novartis, but challenging. The demand for people with these chops runs high across the biomedical, tech and financial sectors. Those who do choose to follow the data science path into biotech research tend to have diverse interests and are passionate about the impact they’re going to have on humanity.

Big Data is pushing the technology of our everyday lives to frontiers that not so long ago were only imagined in science fiction films. We were once thrilled by tech’s ability to give us directions to a concert venue (MapQuest). Then it gave us a better shot at arriving in time for the opening act (Google Maps). Now it can suggest the best route to the show based on real time traffic patterns (Waze). And what’s that coming down the road? A driverless car? Where technology once thrilled us by solving discrete needs, it’s now connecting vast networks of information to blow our minds. All thanks to computer scientists slicing through Big Data with their algorithmic light sabers.

Big Data is leading a similar shift in medicine, from discrete organs to interconnected protein networks. Think about it: When someone we know is diagnosed with cancer, the first question we ask is usually, “What kind?” And the answer is a discrete location: breast cancer, prostate cancer, lung cancer.

What Big Data is revealing is that naming our cancers after their host organs makes about as much sense as naming a musical instrument after the person who’s playing it.

As it turns out, cancers differentiate themselves not by their addresses, but by the genetic mutations that spawn them—HER2+ cancer, TSC1 cancer—and other molecular irregularities. These mutations can crop up in tumors originating from multiple organs in the body.

Well, that sure changes things. It’s like we need to create a whole new cancer vocabulary.

This understanding of interconnected protein networks is shedding new light on disease. While many biotech researchers are focused on discovering treatments that will target individual mutated genes, the teams at NIBR, Dr. Winckler’s included, are taking what they refer to as a Pathways approach. They are zooming in even deeper to pinpoint the exact communication pathways within each cell where a mutated gene has caused crucial, operational signals to run amok. When cancer is involved, Novartis has discovered that killing the molecular messenger is perfectly warranted.

Picture the traffic in a big city during rush hour. You can't solve gridlocks by altering traffic patterns in one main intersection. That may just cause the traffic to flow in unexpected, more complicated ways. Traffic is made up of dynamic, interconnected networks and so are we. To solve a problem in our cells' signaling pathways, scientists must consider all the intersections that play critical roles across and among protein networks.

This targeted strategy of cutting the signal off at the pass could effectively limit the assault of traditional chemo and radiation therapies, and their harsh side effects.

Imagine that. Maybe someday we won’t have to fear the treatment as much as the disease.

Chemo treatments tend to damage both cancerous and healthy tissues. The type of treatment Winckler and her team are after has a precise target and the potential to limit the damage to healthy cells. “That allows us to give targeted therapies at much higher and more effective doses. So you can get more of the right kind of medicine.” Which is to say, the treatment may have a higher likelihood of working.

The full symphony—the termination of terminal cancer—is still entwined in the clamor and clatter of the growing sea of data contained in our bodies.

Scattered refrains and crescendos emerge, pointing researchers in new, promising directions. Highly personalized treatments are evoked by the unique melodies of our data, and a sharper focus promises a more harmonious path to our healing. But we’re still in the first movement of this new scientific era. One in which the audience brings as much to the performance as the musicians. And the maestros’ batons are algorithms.

Looking toward the future with their unique Pathways view, the discovery teams at Novartis are wrangling the Big Data of human genetic information, searching for the hidden melodic bar in an epic, discordant symphony. As the cost of genetic testing continues to fall, what happens when every newborn is sent home from the hospital with a complimentary genome sequence? What will that symphony sound like? And what beautiful music will the new artists coax out of it?