A computational biologist at the University of Tennessee at Chattanooga has just cleared the first hurdle in a global race of experts looking for solutions to challenges to come as a result of older people outnumbering the young by 2030.
The race is the U.S. National Academy of Medicine’s (NAM) Healthy Longevity Global Grand Challenge, and a research proposal by Hong Qin, a UTC professor of computer science, has been selected to advance to the second of three rounds of review. Qin proposed the use of artificial intelligence in searching for a common mechanism in the so-called “biological clock” responsible for how humans age.
Titled, “Uncovering molecular mechanisms of aging clocks with interpretable deep learning,” Qin’s proposal earned him a $50,000 cash award for its selection to advance from the initial, “catalyst” phase, to round 2, the “accelerator” phase of the competition. He will now begin testing his proposal toward documenting results in preparation for his round 2 entry in 2023. It’s the latest work in what Qin said is almost 18 years of studying aging.
“This is an ambitious project. I’ve been studying aging since 2004,” he said. “The second-round review for the NAM competition requires a separate submission and, hopefully, I will have some good preliminary results that show this idea is not just cool, but it actually works.”
That idea is to use create “artificial neural networks” for “deep learning” designed to mimic the human brain’s biological neurons involved in processing information and coming to conclusions or making decisions.
Qin will use neural networks to analyze massive amounts of existing data on biological markers for aging in humans and in mice and several other animal species toward the possibility of detecting a common aging mechanism.
“I’m trying to understand what this biological aging clock is and how it works in humans, a mouse, a fruit fly, even a single-celled organism like a fungus,” Qin said. “This is because, in humans, there is a set of biological markers. In a mouse, there is another set. Although different species have different biomarkers of the aging clock, they most likely reflect a common mechanism which is a black box at the moment. We know these different sets of biomarkers in different species on the surface, and most biologists think there are common mechanisms—or at least overlapping mechanisms—in these aging clocks, even though we do know about them.
“The good thing about deep learning is we don’t have to know what the mechanism is. We can let the artificial intelligence learn—from all the data—what the best model or mechanisms are to explain the data.”
The NAM success also is opening doors. At a recent conference, Qin met the director of a National Institutes of Health (NIH) research program who invited him to submit a proposal to a relevant program. “It has put me on the radar of other research entities,” he said.
Qin credits the UTC SimCenter and its early support and funding for his “basic idea of this line of research” in 2017 with positioning him to succeed in advancing it. Qin also said that SimCenter’s large data storage facility enabled him and his students to analyze the GTEx (genotype-tissue expression) data set for the NAM award challenge. The GTEx data set is made available by NIH through a data usage agreement.
“That data set is so large, I had to use two external hard drives. SimCenter’s large-scale storage made it feasible for us to do some analysis on this extremely large data set,” Qin said. “Without SimCenter’s large-scale data storage, we would not even be able to have handled the data set.”
The SimCenter is a multidisciplinary research hub and the University’s core high-performance computing and storage facility. It is one of 10 such university facilities funded by the Tennessee Higher Education Commission to advance multidisciplinary research in applied computational science and engineering toward growth in research funding, excellence in integrated research and education; and increased stature and economic competitiveness for Tennessee.
SimCenter Director and Computer Science Professor Tony Skjellum said Qin’s work with the facility is an example of how the SimCenter fulfills its purpose.
“That’s part of the mission we have at UTC. We fund scientists like Hong (Qin)—we have peer-reviewed funding, and we have internal startup funding for projects with A.I. and simulation in them,” Skjellum said. “We award peer-reviewed internal funding, and he’s taken that and leveraged it into lots of successes in many different areas of funding and research. We sort of gave him an internal start and he’s taken that to make this latest success.”
Every year, the SimCenter invites research proposals, funding as many as eight or nine with up to $100,000 each for one year of work. Skjellum said the most recent funding awards went to five large-scale proposals and two smaller ones.
Since its launch in October 2019, the NAM Healthy Longevity Global Competition has brought together 11 global collaborators representing more than 50 countries and territories. More on Qin’s catalyst-round win is here.