Meet Jianlin “Jack” Cheng
How do you solve big problems such as fighting disease or decreasing pollution? One protein at a time. Meet Jianlin “Jack” Cheng, a pioneer when it comes to understanding the basic building blocks of life.
Cheng is the Thompson Professor in the Department of Electrical Engineering and Computer Science. He’s spent nearly 20 years solving the mysteries of protein structures. Two years ago, he was named one of the top 100 leaders in the world for using artificial intelligence in drug discovery and advanced healthcare. He and his students continually take top honors at an international protein prediction competition. Earlier this semester, he received the College’s Outstanding Senior Research Faculty Award.
And just hearing him talk about the potential of proteins makes you optimistic and even excited about the future of the world.
From Credit Cards to Molecules
Cheng’s journey to academia has had a few surprising stops along the way.
After earning a bachelor’s in computer science in China, he took a job developing software that could process credits cards. It was the early 1990s; and computers were just starting to become commonplace in the workforce. After five years, Cheng got bored.
So he earned his master’s from Utah State University and became a software engineer for the Sony Corporation in Los Angeles, processing graphics for the movie industry. But not even Hollywood could keep him interested. Cheng wanted to do bigger things.
He found his calling in a lab at the University of California-Irvine, where he earned his PhD. There, he was introduced to bioinformatics and using machine learning to solve health problems.
It’s also where he saw his first protein structure.
“It was beautiful,” he said. “The strands make different shapes, and those structures attracted me intuitively. Then to know that the shape is determined by the protein sequence, and you know there’s something there that can predict the shape, but you don’t know how. That’s the mystery. You know it can be done, but you don’t know how.”
Pioneering a Deep Learning Approach
Proteins are organic molecules that make up all life. They begin as strings of amino acids that fold into three-dimensional shapes. Those shapes then determine how a protein will behave. The spike-like protein found in COVID-19, for instance, acts like a syringe that injects the virus into our cells. Knowing that has allowed scientists to develop vaccines that fight back.
For decades, scientists have been trying to figure out how to predict what shape a protein’s string of amino acids will become. There’s even an international competition to encourage scientists to study the problem.
The Critical Assessment of Structure Prediction (CASP) conference attracts top researchers from academia and industry to showcase various methods of determining how protein sequences will fold. Last year, a Google-owned company came closer than ever to fully solving the problem. (During that same competition, Cheng’s team came in 7th place overall.)
It’s unlikely the company would have been able to do that without the foundation that the CASP community laid over years. Cheng was one of the first to use deep learning in a protein-predicting algorithm, and he shared his developments not only in research journals but also through free software tools available to anyone.
“As engineers, we make products that other people will use to advance their technology and their research,” he said. “We make software available for people to download and use, and that increases the impact of our work.”
Cheng’s latest prediction model is based on a protein’s specific amino acids and how far apart those molecules are from one another. It’s unique in that it provides explainable information to support the prediction.
Many researchers around the world use it to study specific proteins related to cancer and other diseases; crop production; and green energy.
From Cancer to Climate Change
A few years ago, Cheng began tackling climate change by looking into green energy himself. He received a grant from the Department of Energy to study proteins found in green algae in hopes of finding a way to boost production and eventually allow algae to replace fossil fuels.
Green algae grow in water and convert sunlight into energy. It’s especially of interest as a biofuel because it also absorbs carbon dioxide out of the atmosphere. So it would reduce carbon dioxide pollutants not only when used as a biofuel, but also while it’s growing.
Right now, though, the cost of producing green algae for energy is higher than current fossil fuel production.
But that could change if Cheng and collaborators are able to engineer proteins within algae to foster growth. They’re using AI to guide them as to which amino acids need to be altered to make that possible.
“If we can make it more efficient, we can continue to reduce the cost of cultivating green energy and reach a point where it’s cheaper or at the same price as fossil fuel production,” he said. “And that’s when we can reduce climate change and make more sustainable energy.”
Paying it Forward
When he’s not tackling major world problems, Cheng works hard to train the next generation both in the classroom and as an academic advisor.
Students in his courses get the best of both worlds—an understanding of the theories behind computer science and the knowledge of how those theories are used.
“If you only learn theory, you don’t know how it works in the real world, but if you only work on the real-world problems without theory, you don’t have sufficient guidance,” he said. “You have to combine the two.”
He uses class projects to help students develop problem-solving skills and teamwork while building a strong theoretical foundation.
Cheng also knows how important advising is. Just as he found his passion during his PhD studies, he knows he’s helping future generations find theirs. And for some, that has meant going on to become professors themselves.
“If I train students to become professors, then they train their students, you spread the knowledge and skills quickly,” he said. “So you start to develop a high-tech workforce. That’s how the U.S. maintains its competitiveness in the high-tech domain. In order to keep doing that, a research-driven education pipeline must be maintained.”
One Protein at a Time
Cheng doesn’t get overwhelmed at the prospect of trying to cure diseases or solve climate change. They’re enormous problems, but he keeps his eye on the tiny proteins that could become solutions.
In recent years, he’s turned his attention to RuBisCO, an abundant enzyme that accounts for a large portion of proteins found in all plants.
“I didn’t know much about this protein before, but it’s arguably the most important enzyme on Earth” he said. “All plants use it to convert carbon dioxide into biocarbon that can be used by life systems. So without this enzyme, there is no life on earth.”
With such a big role in our existence, there’s no telling what a better understanding of RuBisCO means for human life.
For now, though, Cheng believes it’s key to biofuel production and a cleaner, more energy efficient planet.
“I like to work on fundamental problems like treating cancer and producing green energy,” he said. “Those are ambitious goals. But we work on manageable, specific problems that can be solved in a couple of years. By having a big vision with very significant targets, we can make real progress. I’m very excited about it.”
Want a computer science education that gets you excited about research possibilities? Check out the EECS Department at Mizzou Engineering.