Join MySLAS Social

Gut Reaction: SLAS2019 Keynote Speaker Eran Segal Decodes the Human Digestive Tract to Discover Precision Nutrition

A quest for a personal best resulted in a shift in research focus for Eran Segal, Ph.D., who digs deep into the microbiome of the human gut for clues to optimizing health and preventing and fighting disease by building a better diet.

Eat less fat. Eat fewer carbs. Don’t eat dairy, red meat or wheat. Limit alcohol and caffeine. The list of dietary dos and don’ts goes on forever, and the science behind it is not all that convincing, says Eran Segal, Ph.D., a professor in the Department of Computer Science and Applied Mathematics at the Weizmann Institute of Science (Rehovot, Israel) and a keynote speaker for SLAS2019 (Feb 2-6, Washington, DC, USA).

“What is the best diet for humans? After reading years of research, I wanted to know why there was not a definitive answer,” Segal says. What he discovered is that researchers were asking the wrong questions, “because it assumes that the best diet depends only on the food and not the person eating it.”

Segal and Eran Elinav, M.D., Ph.D., a researcher in Weizmann’s Department of Immunology, decided to shape the exploration of nutrition in a new way. Segal and Elinav explored how differences in genetics, lifestyle and gut bacteria might cause people to respond differently to foods. What if these differences explain why some diets work for some people, but not for others? What if nutritional needs could be tailored to each person’s unique makeup?

The pair’s results suggest that personalized diets may successfully modify an elevated post-meal blood glucose level and its metabolic consequences. After publishing their work in a scientific journal, the duo went on to publish a self-help book, The Personalized Diet: The Pioneering Program to Lose Weight and Prevent Disease, to help the average reader understand the science behind the research, provides tools to create an individualized diet and lifestyle plan (based on glucose reactions) and put the reader on a path to losing weight, feeling good and preventing disease.

“What makes our research unique is that we’re trying to look at the individual, not just groups of people,” says Segal. “We take a data-driven approach to collect relevant and high-quality data from people with the goal of building a predictive model that takes the parameters of the individual into account. This is where personalized, precision medicine begins, within each individual’s make up.”

Machine Learning Guides Precision Nutrition

The initial spark that nudged Segal toward nutritional research was a desire to improve his marathon time eight years ago. “I tried to think of all the ways in which I could optimize my physical ability to achieve a time of less than three hours, and one of the areas I decided to examine was nutrition,” says Segal. “As I read the literature, I was disappointed by the nutritional advice we follow today and what poor science, in my opinion, on which it is based.”

Research released in 2012 that delved into the microbiome of the human gut captured Segal’s attention. “It’s a system of approximately 100 trillion microbes and up to 1,000 different microbial species, which is 10 times more cells than we have in our entire body,” he says.

“The research reveals amazing and mind-blowing experiments and results in the microbiome,” which includes scientists transferring gut bacteria from obese people into mice.“ After a few weeks, the mice copied the phenotype of the donor, developing obesity even though they were receiving a lean diet,” Segal reports. Another study he read reveals insulin resistance associated with certain communities of gut microbes in humans. Yet another shows a strong association between type 2 diabetes and a certain fingerprint of gut microbes. Segal knew immediately that this was an area of developing importance in health and nutritional research that he wanted to expand on in his research with Elinav.

To launch their own studies, Segal and Elinav searched for a metric of health nutrition to guide their research, one that they could easily and accurately measure across many people. They focused on blood glucose levels, which are associated with many health complications, such as type 2 diabetes, heart disease and stroke. Higher levels also encourage the body to secrete more insulin, which results in more sugar stored as fat, and thus weight gain.

The researchers continuously monitored week-long glucose levels in an 800-person test group, measuring responses to 46,898 different meals and found high variability in the response to identical meals. Some had strong glucose reactions to diet-busting culprits such as ice cream or chocolate while others didn’t. Some people could eat bread and rice without effect, others could not. This discovery reveals that universal dietary recommendations may have limited utility.

The next step was to accurately predict personalized post-meal glycemic responses. To achieve this, Segal used the results of the 800-person cohort to develop a machine-learning algorithm that integrates blood parameters, dietary habits, anthropometrics, physical activity and the microbiome of the digestive tract.

The team validated the algorithm using an independent 100-person cohort. A blinded randomized controlled dietary intervention based on this algorithm resulted in significantly lowering post-meal responses and consistent alterations to the gut microbiota configuration. The data indicates that personalized diets may successfully control post-meal blood glucose levels and their metabolic impact.

“What started out as a personal interest developed into an opportunity to do some real, basic quantitative research that hasn’t been done well in the field of nutrition,” says Segal. “Shifting the focus of my studies to nutrition was very natural. Studying the microbiome is a modified DNA sequence analysis, which is what I had been engaged in for the past 20 years.”

Segal’s multi-disciplinary lab, which consists of computational biologists and scientists studying the microbiome, nutrition, genetics and gene regulation in health and disease, continues to focus on decoding the signatures of disease at all levels of omic data sets. His aim is to develop personalized nutrition and medicine using machine learning, computational biology, probabilistic modeling and analysis of heterogeneous genomic and clinical data.

“It’s more of a holistic view, and we are developing computational models of people at various stages and studying the human biome in ways that technology is only now allowing us to do,” says Segal. "All of these opportunities are new. The ability to be at the forefront and be a pioneer in these types of studies is exciting. Everywhere we look is an opportunity to study some aspect of nutrition for the first time and to examine fundamental questions in human health. It’s not surprising that we’re identifying a lot of new discoveries at different levels in the data sets.”

Segal hopes that his research leads to a new understanding of the mechanisms and prevention of disease and discovering the role that the microbiome plays in all of it. “We hope to better identify the trajectory that people are on and be able to intervene and change the course of people’s lives for the better,” says Segal. “Research that has the potential of transforming the way we think about and monitor our health is what excites me the most.”

Biological Discovery through Big Data

A computer scientist and mathematician by training, Segal became a self-taught programmer at the age of nine when he got his first computer – a Commodore 64. These early interests eventually led him to an undergraduate degree in computer science from Tel Aviv University (Tel Aviv, Israel) and then to Stanford University (Stanford, CA, USA), where he earned doctoral degrees in computer science and genetics.

“I started in 1999, about a year before the Human Genome Project was published,” Segal says, adding that around the same time, Stanford’s Pat Brown, Ph.D., (who introduced Impossible Food in 2017) invented the microarray. Segal, who worked in the area of machine learning, describes it as an exciting time of turning points for computation and biology.

“A lot of the buzz words that one hears now – the elements that are changing how we do artificial intelligence and machine learning, such as big data, algorithms, Bayesian networks and deep learning – were developed at that time. These are the types of models on which I worked and programmed through my Ph.D.,” Segal continues. "We moved from qualitative to quantitative science. I was lucky to be at the transition point. It seemed to me that something big was there, but it was just getting started.”

After completing his Ph.D., Segal began an independent fellowship at Rockefeller University (New York, NY, USA). “I had the freedom to think about what I wanted to do because I was not assigned to any particular lab. I developed a research agenda around epigenetics,” he explains, adding that he was lucky to meet colleague Jonathan Widom there.

In 2005, he returned to Israel to join the Weizmann Institute of Science and open his lab. “The lab started out computational, but after two years I realized that no one was producing all of the data that we wanted and needed for our work. I decided to jump ahead and open a real wet lab, even though I didn’t have any experience with experimentation,” Segal says. The team started small and in time developed many molecular technologies, such as massively parallel reporter assays – the ability to synthesize tens of thousands of DNA sequences, insert them into either yeast or human cells and then very accurately measure the expression driven by each sequence or the effect that these sequences had on cellular growth. At one point, Segal notes, there were more staff members on the biological side than the computational side.

“Our 35-member lab staff now is a mix of experimentation and computation that includes students, postdocs, research associates and technicians. It fluctuates depending on what we want to accomplish,” Segal continues. “We’re running large scale clinical trials with human subjects, so we’ve been joined by research coordinators and dieticians. We have a lot of collaborations, but we also have the in-house skill set to run diverse research projects and control all aspects of them. Our research can have a profound impact on people’s lives all across metabolic disease, heart disease, multiple sclerosis, cancers, type 1 diabetes and even the development of the infant microbiome, particularly in premature birth.”

Segal acknowledges that it’s hard to step away from this all-consuming work – particularly because his launch into his current area of research was so personal. “My work doesn’t really feel like work,” he says. “When you add smartphones that allow you to work from anywhere at any time, sometimes it’s not necessarily good.” In spite of this, Segal adds that he still runs most days and enjoys swimming, cycling and hiking with his wife – a clinical dietician – and their three children.

SLAS Keynote Presentation

“SLAS2019 is a unique opportunity for me to step outside my specific field and the interactions I have with colleagues at work and reach out to the Society’s large, diverse community of researchers,” Segal says. “I am excited by this opportunity to present our studies, because the implications and usefulness translate to many people.” 

His presentation, “Personalized Medicine Approaches Based on the Gut Microbiome,” provides an overview of his lab’s many research projects that examine broad questions. “I plan to reveal some of the trials that we’re doing and speak about some new, unpublished data that we’ll have at that time,” Segal says.

He hopes to inspire in scientists a drive to find what interests them, using his own journey as an example. “When you find what interests you, it should be big enough to motivate you and help you push through challenges. Having a driving interest helps you reach results,” says Segal, who also strongly encourages researchers to seek a mentor or two along their career paths. “Having someone to guide you through the process is important. We don’t talk about this much in science but getting the right support for doing research is key.”

Segal acknowledges the challenges of research. “It is hard and can be frustrating sometimes because things don’t work out, they take time, they aren’t always structured. Much of the work lies in the details,” says Segal, who advocates a balance between tenacity and flexibility. “Do not become too stubborn about one particular approach or strategy. But at the same time, don’t walk away from what really interests you.”

Segal concludes, “I have done a few pivots in terms of what I studied, but I made those changes not because one area wasn’t successful, but because something else drew my interest and I decided to follow it. I think one great aspect of being an academic researcher is the freedom to choose your work. I think many people don’t take as much advantage of this as they could.”

October 15, 2018