Computational Biology, Systems Biology
Massachusetts Institute of Technology
Dr. Regev is also an associate professor of biology at the Massachusetts Institute of Technology and a core member of the Broad Institute of MIT and Harvard University. She was an HHMI early career scientist from 2009 to 2014.
Accepted into a small, highly selective program at Tel Aviv University, Aviv Regev was drawn to computer science, math, and physics. But when she added classes in genetics and evolution to her freshman course load, she was hooked. "I was fascinated by how the cells process information and act like little computers," Regev recalls. "I still loved the computation, but my heart became about the biology."
For her PhD research, Regev applied an approach from computer science called pi calculus to networks of biological molecules. Concurrently, she established and led a bioinformatics team at Quark Biotech, a functional genomics company. Soon after, Harvard recruited her for a Bauer Fellowship, which gave her five years of complete freedom and no need to apply for grants. "That was like heaven," Regev says. She created pioneering algorithms such as Module Networks, computational tools for reconstructing networks, and Synergy, which reveals how circuits are rewired during evolution. "I was working on trying to understand circuitry in cells, both in the short term as cells are stimulated in some way and in a long-term process like evolution," Regev explains. Those algorithms are now being used in labs around the world to analyze gene expression and other genomic information.
Regev's computational tools were making sense of the mass of data coming from biologists' labs, but she wanted more. "I got this itch to do certain experiments," she says. Her idea: use the computational wizardry to identify potential circuits and regulatory pathways, and then test and manipulate those presumed circuits at the lab bench in actual cells. "We know how to develop algorithms that make good predictions, but we know they are just predictions," she explains. "So we have to go back and test experimentally. Just doing one or the other will never solve the problem."
After setting up her own lab at the Massachusetts Institute of Technology and the Broad Institute in 2006, Regev began to apply her algorithms and test them in yeast cells, dendritic cells (the watchdogs of the immune system), and immune cells called T helper 17 (Th17) cells, which are implicated in autoimmune disorders, such as multiple sclerosis (MS) and rheumatoid arthritis.
The approach has been extremely fruitful. In one recent study, Regev's team partnered with immunologist Vijay Kuchroo and his team at Harvard Medical School to explore the complex circuitry that connects a key receptor on the surface of Th17 cells to gene transcription. Mutations in the receptor are associated with autoimmune disease.
The algorithm identified probable paths for a signal to travel through what Regev calls a "hairball of proteins." The computational model then ranked the importance of each node in the protein network. "The top-ranking node was a protein that no one had related to these Th17 cells before," says Regev. It proved to be biologically crucial, offering new possibilities for treating human disorders. Knockout mice without this protein didn't get autoimmune disease, for instance. Moreover, the protein is activated by salt. Feed extra salt to mice that are susceptible to MS and they develop MS. "Because it's not human, the algorithm has the opportunity to see things you wouldn't see. Then once it shows it to you, you can use your biological knowledge to find out if it's true."
In another recent study, Regev and her colleagues analyzed the whole course of differentiation of Th17 cells. They used nanowires developed by physicist Hongkun Park at Harvard University to insert genes into the cells to perturb the process. They discovered a collection of factors that activate the differentiation process and another set that inhibit it, creating a system in perfect tension, she says. It makes biological sense. "The balancing act is the essence of the differentiation process," she explains. "If there are too many Th17 cells, you get inflammation; too few, and there's an increased risk of infection."
Regev's group now is developing microfluidics, or lab-on-a-chip, techniques to chart gene expression in hundreds—and eventually many thousands—of single cells every day. She's also directing a new project called the Cell Observatory at the Broad Institute to identify and map all the circuits in human cells. "I passionately feel the time of this problem has come," she says. "We have the opportunity to answer the fundamental questions of how the cell decides to do what it does and how circuit malfunctions lead to diseases." She hopes this work will open the door to new treatments.