Full Charge Ahead
There are two types of people: those who think cell phone batteries should hold more charge, and those—like Julia Greer—who think cell phone batteries should hold more charge and decide to do something about it.
Caltech scientists and engineers are translating inquiries into insights and ideas into inventions.
Take a look.
There are two types of people: those who think cell phone batteries should hold more charge, and those—like Julia Greer—who think cell phone batteries should hold more charge and decide to do something about it.
One day, people with paraplegia may be able to stand and walk, thanks to a treatment that Caltech graduate student Ellen Feldman is helping develop.
Mitch Guttman fashioned himself a biologist as well as a builder of scientific tools, methods, and algorithms when he arrived at Caltech in 2013. Others, the assistant professor of biology thought, would translate his fundamental science into more effective treatments for patients.
Today, he has a different perspective.
In her third year at Caltech, biological engineering undergraduate Anusha Nathan took her education to a new place—5,000 miles away. She enrolled in Caltech’s exchange program with École Polytechnique, one of France’s top universities in bioengineering, biology, and health sciences.
We translated her 12 weeks in France into a set of lessons—about science, language, and life.
We gave five Caltech postdoctoral scholars and graduate students a challenge: Using the simplest terms possible, tell us about a concept related to your research. Here’s what they said.
I’m Stewart Mallory. I’m going to define self-propelled nanoparticles.
These are synthetic particles that you can make in a lab and they have their own source of energy. So they have their own engine, and they can be propelled around.
So you can think of the Magic School Bus. In the Magic School Bus, the kids get in the school bus, the school bus shrinks, and it goes to all these places on the microscale.
These particles, they’d be small enough to actually swim inside of your blood vessels. And so you could inject these particles into a site, and they could swim up to where you have this clogged artery, and they could actually unclog it for you.
So I use computer modeling and simulation to really understand how these things behave. Every day when you’re doing these simulations—you know, it’s like there’s the potential that you’re the first person to really look at this behavior and see these things.
Kind of why I get up in the morning.
My name is Ezgi Kunttas. Let me define collective cell migration.
You can imagine in nature a lot of organisms migrate. For example, birds migrate, monarch butterflies migrate. And, believe it or not, in our bodies also, collective cell migration happens where multiple cells migrate together. This happens both during normal development, but also in a disease state such as cancer invasion. I’m interested in how cells communicate with each other while they are migrating.
When I was watching a specific movie that I was doing in zebrafish migration, the cell in the front somehow got detached from the rest of the cells. And they started migrating, and then it just stopped—realized that something was not right. So then it actually turned around and [went] back to the rest of the group. They waited for that leader cell to come back. And then the migration continued.
It’s a fascinating process to watch under the microscope.
My name is Chujun Lin. I study human face perception.
Imagine you come across a profile photo of a stranger. Do you ever wonder: “What is he like? Is he shy? Is he trustworthy?” Do our impressions convey our stereotypes and biases?
These questions have fascinated psychologists for a long time. Compared to the literature, the approach at Caltech is more quantitative and mechanistic. For example, in one study we manipulated the politicians’ photos. What we found was that they reported that the same politician looks more corruptible when their face was made fatter and less corruptible when their face was made slimmer.
Here at Caltech, we are paying particular efforts to make our studies transparent and replicatable. Understanding the stereotypes will help people to make more rational decisions. For example, when people decide which candidate to vote for or when people are browsing the photos on dating websites.
My name is Alejandro Robinson Cortes. I’m going to define the economics of matching in foster care.
A striking fact in foster care is that a majority of the children go through more than one foster home while they are in foster care. This is detrimental for children’s development.
So, foster care you can think of as a market, even though it’s not that obvious. Children that need to be taken care of are the demand. And homes—private families, foster homes that are willing to take care of them—are the supply.
One of the goals of my research is to identify in which ways we can match children in a certain way in which we can minimize disruptions. From the data, I’m going to be able to obtain a set of characteristics that tells us when you match this child with this foster home, this is less likely to be disruptive.
We’ve learned that economic tools help us to understand a really wide range of phenomena, something that was not known 50 or 70 years ago. It’s a new field of inquiry.
I’m Nicole Yunger Halpern and I’m going to define quantum information in quantum cognition.
Most physicists will not say anything about quantum theory together with cognition, except maybe over a beer in the evening.
Brains are warm, wet, and large. In these conditions, quantum entanglement—the very strong correlations between quantum particles—die.
Matthew Fisher is a physicist who proposed a mechanism by which quantum entanglement—these correlations—might affect neuron firing. Entanglement might be shared by spins, quantum properties of phosphorus nuclei. Being in small molecules could shield these spins from the warmth and the wet. These spins would be carrying quantum information.
So what computations might be happening in us? We don’t know if quantum information affects cognition. We need data. And if any support this proposal, then our view of biochemistry and ourselves might have to change.
Nanorobots to the Rescue
Learn about self-propelled nanoparticles from Stewart Mallory, an AGEP postdoctoral scholar in chemical engineering.
Birds and Butterflies Do It ... and So Do Human Cells
At any moment, millions of our cells are on the move, orchestrating immune responses or other benefits. Ezgi Kunttas, a postdoctoral scholar in biology and biological engineering, discusses collective cell migration.
Before You Vote or Go on a Dating Website, Think of This
Social science graduate student Chujun Lin shares her research on human face perception.
How Economics Can Help Children
Alejandro Robinson Cortes, an economics graduate student, describes how he is applying matching market theories to the U.S. foster care system.
Quantum Theory and Our Warm, Wet, Large Brains
Most people have a hard time grasping the concept of quantum information. What does it mean to pair the term with quantum cognition? Physics graduate student Nicole Yunger Halpern explains.
There are two types of people: those who think cell phone batteries should hold more charge, and those—like Julia Greer—who think cell phone batteries should hold more charge and decide to do something about it.
The rechargeable lithium-ion battery currently powering your cell phone or laptop produces an electrical current through chemical reactions at two terminals: the cathode and the anode. The Greer lab started their battery overhaul at the anode.
The anode is where lithium atoms are released and where the electrons that power your device are generated. It is commonly made of graphite, a form of carbon, which is chosen for its high durability and satisfactory ability to absorb lithium.
But silicon, one of Earth’s most abundant elements, is at least five times better than carbon at storing lithium. That would make a battery that lasts much longer. Regular silicon wouldn’t make a good anode, though, because when it takes on lithium, it expands to the point where it breaks down. Until now.
But before it reached this stage of development, how did Greer translate her idea into an intellectual pursuit that then blossomed into proof-of-concept data? With seed funding, of course.
After she came up with an idea that just might revolutionize the way batteries are constructed, Greer applied for a grant through Caltech’s Rothenberg Innovation Initiative (RI2).
That grant made it possible for her to begin her research, working with grad student Xiaoxing Xia and postdoctoral scholar Heng Yang. Later, in 2016, she applied for and received support from the Samsung Advanced Institute of Technology, which enabled her team to carry on the investigations.
Greer’s lab has found a way to let silicon “breathe” rather than break. The scientists successfully engineered a 3-D microscale lattice (picture a dust mite-sized Eiffel Tower) and coated its beams with an ultra-thin layer of silicon. Now, when the silicon absorbs lithium and increases in volume, it has room to expand into the open spaces of the lattice architecture.
The next phase in the team’s ambitious quest is engineering a safer electrolyte—one that won’t degrade or overheat with use.
The electrolyte is the part of a battery where lithium ions travel between the cathode and anode. In today’s batteries, it is a flammable organic liquid that, infamously, on occasion gets too hot and makes cell phones catch fire.
Greer’s team aims to replace that liquid with an ultra-thin layer of solid electrolyte on their micro-beams, forming an interconnected, highly conductive 3-D highway. These beams can store and release large amounts of energy much more efficiently than today’s state-of-the-art rechargeable batteries.
In 2017, Caltech’s Office of Technology Transfer and Corporate Partnerships helped facilitate a partnership between Greer and a venture capital firm that invests in groundbreaking science. The company endeavors to scale up production of Greer’s nanoscale designs to accelerate the development of materials with improved optical, electrical, mechanical, and thermal properties.
Think fire-resistant fabric, lightweight concrete, and shatterproof glass.
One day, people with paraplegia may be able to stand and walk, thanks to a treatment that Caltech graduate student Ellen Feldman is helping develop.
Ellen Feldman. The figure on the right draws on her spinal model, with neural matter (blue) cushioned in cerebrospinal fluid (yellow), encircled in fat (green), and enveloped by vertebral bone (orange).
The young computer scientist, who in 2017 won one of the first Amazon Machine Learning Research Awards, has joined a collaboration involving Caltech, other universities, and the Frazier Rehab Institute. The project has brought researchers and doctors together to develop a new therapy for paralysis. They start by implanting electrode arrays in the backs of paralyzed patients. In effect, the electrodes can send signals to the patients’ spinal cords.
“People who have suffered a traumatic injury to the spinal cord can’t stand or walk,” Feldman says. “But delivering electrical stimulation to the spinal cord has been shown to help. This treatment restarts networks of neurons that have survived below the site of the injury.”
Spinal-cord injuries vary widely in terms of the location, type, and extent of damage. Thus, an electrode stimulation pattern that works for one patient may not help another. That means that doctors have to customize the array settings for each individual.
Since arrays contain at least 16 electrodes, each with three settings, clinicians have more than 43 million treatment options for any given patient. “And some arrays have 32 electrodes,” Feldman says, “which makes the search even harder.”
If this treatment is to work well, doctors need a systematic way to decide which pattern to apply. That is a problem that machine learning could help solve, thinks Feldman’s adviser, Joel Burdick.
But applying machine learning in this way is unconventional. Before he could win federal grants to advance the idea, Burdick needed to prove its potential. So, with seed funding from the Christopher and Dana Reeve Foundation, he recruited machine-learning experts.
One of those experts is Yisong Yue. Yue had just arrived at Caltech when he heard about Burdick’s idea some four years ago. He felt intrigued but skeptical. “Normally, I deal with companies such as Facebook or Google, where there is a tremendous amount of data to work with,” Yue says. “Creating algorithms for a situation where the amount of data being generated is much lower presented some unique challenges.”
On a conceptual level, the algorithms work like the personalized recommender systems that Netflix and Amazon use. First, researchers test one stimulation pattern. Then, algorithms incorporate feedback from doctors and patients to select the next pattern.
“We want to be able to provide customized treatments that take into account a particular person’s physiology, injury, and even the location where the device is surgically implanted,” Yue says.
After Yue and Sui collect experimental data, Feldman’s work begins. She employs machine learning differently, to gain a big-picture view and predictive capabilities. To start, she created a digital wireframe representation of the nerves, bones, muscles, and fat that make up the human spine. Now, she feeds the experimental data into this model.
By learning from the results, she can predict how patients’ damaged spinal cords will respond to untested stimulation patterns.
Image: Move the slider arrows left and right to see how one of Feldman’s simulations maps the voltage and electric field measurements resulting from a spinal stimulation.
As they gain predictive capacity, her algorithms will better identify the patterns most likely to stimulate the surviving neurons. By homing in on how a given patient will respond to a given treatment, her model can eliminate guesswork.
Now, Feldman is connecting with neuroscientists to share and refine her ideas. Because of her Amazon funding, she could afford to travel to a Society for Neuroscience annual meeting in Washington, D.C. There, she spent a week learning about new neuroscience research, including findings relevant to her work.
“It was an amazing experience,” Feldman says. “I saw an interesting study comparing how spinal-cord stimulation affects humans and rats. That study could give us better understanding of specific spinal circuits being stimulated.”
Feldman values the opportunity to translate her training and skills in computer science and machine learning into medicine. Looking back, she traces her interest in the intersection of these disciplines to an undergraduate summer internship at Johns Hopkins University. While there, she worked with surgeons in the medical school to help develop robots that could aid them during operations. “I thought that it was so inspiring to be able to work on a tool that would eventually help doctors and patients.”
In the future, she and her colleagues hope to test her model-generated patterns in real patients—helping them stand and perhaps even walk again with the aid of robotic exoskeletons.
Mitch Guttman fashioned himself a biologist as well as a builder of scientific tools, methods, and algorithms when he arrived at Caltech in 2013. Others, the assistant professor of biology thought, would translate his fundamental science into more effective treatments for patients.
Today, he has a different perspective.
Guttman and members of his lab continue to reveal basic insights into human development. And they invent tools that open new windows onto these essential processes. But now the team aims to use the knowledge they gain to devise entirely new ways to treat genetic diseases.
Guttman has dedicated his career to deciphering the functions of long noncoding RNAs (lncRNAs), a class of genes he helped discover when he was a graduate student. There are as many lncRNA genes as there are protein-coding genes, but lncRNAs are not nearly as well understood.
He applied for some grants to develop specific tools to study lncRNAs, but not all funders immediately took to his ideas. Some of the rejection letters stated that his proposals were too risky.
It was private philanthropy—including his appointment as a Heritage Medical Research Institute (HMRI) Investigator at Caltech, which provides a research stipend—that enabled Guttman to persist.
“I don’t know if the granting agencies thought I was some dumb young kid,” Guttman says. “But what makes Caltech and Dick Merkin [the Caltech trustee who founded HMRI] special is that they understand and appreciate how great science happens: You allow smart people to take risks and try new things. Will these ideas always work out? Maybe not. But if we can make it work, we can solve something important.”
Using his research funds, Guttman sought to construct new methods for unraveling the mysteries of lncRNAs.
He and his lab have devoted part of their effort to Xist, a specific lncRNA that is responsible for the silencing of the second X chromosome in females. (Nearly all females have two X chromosomes, but only one of these chromosomes is active.)
The team has created a series of important tools—with humorous names such as RAP, Gin, and SPRITE—that have revealed new facts about the function and structure of RNA. For the first time, scientists pinpointed the exact proteins that interact with Xist and confirmed that RNA can change the structure of DNA.
Video demonstration of the Gin (Genome Interface) software that enabled Guttman and his team to visualize the three-dimensional structure of genes
With this new information, the researchers uncovered a biological pathway—a series of actions—that triggers the silencing of the X chromosome. It was not long before Guttman saw applications outside of the lab.
“In most genetic diseases, there is no functional copy of the disrupted gene,” Guttman says. “But that’s not the case for females who have X-linked diseases. They actually have a functional copy of DNA encoded in their genome—it’s just not being expressed.”
With this insight, Guttman now had a strategy: Activate the copy on the second X chromosome.
Collaborating with Caltech chemistry professor Brian Stoltz, Guttman and his team are focusing on this process. Their goal is to develop targeted therapeutics for X-linked disorders such as Rett Syndrome, a rare but devastating condition that originates from a mutated gene on the X chromosome and severely impairs patients’ ability to walk, talk, and use their hands.
Now, thanks to private funders—a connection made possible by Caltech’s Office of Technology Transfer and Corporate Partnerships—the researchers expect to enter phase-one clinical trials in the near future.
If successful, this approach could help people living with Rett Syndrome and other X-linked genetic diseases, including certain types of cancer.
Guttman’s and Stoltz’s work exemplifies the potential of fundamental discovery to inspire breakthroughs in the clinic. With its new Translational Sciences and Technology for Health initiative, Caltech aims to do even more to leverage expertise across the basic sciences and engineering to improve lives.
In her third year at Caltech, biological engineering undergraduate Anusha Nathan took her education to a new place—5,000 miles away. She enrolled in Caltech’s exchange program with École Polytechnique, one of France’s top universities in bioengineering, biology, and health sciences.
We translated her 12 weeks in France into a set of lessons—about science, language, and life.
Although conversant in French, Nathan had never had much opportunity to talk about biomedical devices or the mathematical modeling of medical phenomena.
So the term before she went to École Polytechnique, she took L175, French Conversation. The course, led by lecturer Christiane Orcel, enlists French-speaking researchers from Caltech and JPL to present about their research in the language.
“The only thing I’d done up till then was grammar rules and cultural communication,” Nathan says. “L175 gave me the foundation for getting into the mindset of thinking about science in French.”
They pepper day-to-day speech: the nervous “umm,” the stalling “like,” the equivocating “sort of.” Not typically taught in formal language training, these utterings also don’t translate directly between French and English—and presented an early stumbling block for Nathan.
“I had wordreference.com open, looking up ‘enfin,’” she says. “It’s defined as ‘finally,’ and some phrases don’t make sense without it, so I was writing it down in my notes. I found out later it’s used to mean ‘okay,’ ‘so,’ or ‘just.’ It just flew over my head at first!”
One class about quantitative imaging in biology at École Polytechnique required some complicated coordination. The discussion was in French, but the programming was in Python, which is based in English.
“It was the most trippy thing, trying to listen to the professor and students, figure out what they were saying in French, write it in English, then repeat it back in French,” Nathan says. “It took some time juggling the back and forth, but eventually I was like, ‘Got it!’”
Participating in a sport is mandatory at École Polytechnique, so Nathan joined the rowing team and the soccer club.
“I came back maintaining that perspective on life, thinking about how I was going to balance my classes better with all of the other things I like to do,” she says. Nathan has been inspired to take up regular weightlifting sessions and to redouble her commitment to dancing with Aarya, an Indian classical dance team she cofounded at Caltech. Meanwhile, she also is working on her senior thesis with off-campus researchers at City of Hope.
A term learning in Europe also meant trying new foods, meeting new and different friends, and visiting new places such as the United Kingdom, Germany, and Croatia during time off. That spirit of novelty also influenced Nathan to make what was for her an unconventional choice: She danced with the cheerleading team.
Nathan says: “Now I’m open to trying a bunch of new things here, whether it’s a weekend trip or a class that’s outside of my comfort zone. Even if I’m not sure I’m going to understand completely, that’s okay. I’ll still get something from it.”
At École Polytechnique, Nathan stayed connected with stateside classmates and family through social media, video chat, and even visits. She treasured those chances to share her overseas experience. “Recounting your stories cements those memories that are going to stay with you forever,” she says.
Back at home, she remains close with her fellow exchange students—even though they’re far away. She plans to make it to a reunion across the pond one day. “They’re some of my best friends,” she says.
“The exchange program was one of the best experiences of my life,” says Nathan, who received scholarship support from the ARCS Foundation and the Henry Herbert Smythe Trust.
Anusha Nathan (left) with her sister Eisha Nathan
Receive The Caltech Effect and the Caltech magazine newsletter