Illinois ECE PhD student improves bioinformatics software to better analyze human DNA
Human beings have around 10 trillion cells in their bodies, resulting in about 10 billion miles of DNA - almost twice the diameter of the solar system. Illinois ECE PhD student Zachary Daniel Stephens is changing the way medical experts examine and analyze this vast amount of DNA in the human body, ultimately contributing to the trajectory toward individualized medicine.
Stephens splits his time between the University of Illinois at Urbana-Champaign and Rochester, Minn., where he is a research affiliate at the Mayo Clinic working in clinical bioinformatics, a field that connects computing, biology, and medicine. Stephens is applying the skills he learned in the research group of his adviser and Illinois ECE Professor, Ravishankar K Iyer, who is driving an initiative between Illinois and Mayo in computational genomics.
“The analysis of biological sequence data is central to advancing personalized medicine,” says Iyer, the George and Ann Fisher Distinguished Professor of Engineering.
“Zachary develops bioinformatics methods which leverage 'long-read' sequencing data to enhance our understanding of disease etiology in ways that conventional approaches cannot.”
In combining analytics software with modern medical science, bioinformatics allows medical professionals to dig deeper into human genetic code. In Stephens’ area, this work involves scouring human DNA and RNA to find aberrations which may be the underlying causes of various diseases. These mutations, or ‘variants,’ play a major role in human development and disease.
“There’s something up at Mayo called the Diagnostic Odyssey, which involves patients with rare and mysterious diseases,” says Stephens, who is pursuing his doctorate in electrical and computer engineering. “While bioinformatics alone might not completely crack the case, much of my work aims to provide researchers with additional variants that might not have been identified with existing methods. This is an additional 'haystack' where they might eventually find a needle.”
According to Stephens, there are many questions one could ask when examining a person’s genome or transcriptome, and it is not always clear how to connect these observations to what a patient is experiencing. The field of bioinformatics provides an elaborate arsenal of methods to use for each step in the process of developing an actionable outcome - such as a diagnosis. He says the current software is good at what it does, but isn’t entirely comprehensive.
“There are certain things that aren’t addressed as sensitively as they could be,” says Stephens. “That’s why I developed algorithms to bridge the gaps that are left by existing technology.”
While he has now worked in bioinformatics for half a decade, his background is in control theory and signal processing.
“Biology wasn’t my background at all,” says Stephens. “In meetings with clinicians, I would keep a running list of every term they used which I didn’t know, so that later on I could read up on it and show up to the next meeting a little less clueless. If you do that enough times, you’ll eventually pick up enough to understand what the doctors are talking about.”
It was this approach to a new field that lead Stephens’ evolution at Mayo Clinic. While he now works closely with medical staff and his focus is bioinformatics, he started off working closer to what he knew.
“It went from being very much about computer science and IT to actually being ‘on the ground’ clinical and bioinformatics work,” says Stephens. “I think a lot of that has to do with what they expect from engineering students versus how much they trust you to work with data that directly impacts patients.”
Stephens says that while his field is important, it’s only part of a much larger system that all works together to develop innovative solutions to medical problems.
“All the bits and pieces have a thousand variables,” says Stephens. “All of my work goes on to influence other pieces of the larger medical pipeline.” Iyer is also affiliated with the CSL.
Check out the original article on the CSL site.