UI researchers develop faster, cheaper DNA sequencing
Sequencing a genome is still laborious and expensive, the kind of thing you do for the human race but not for a human – not routinely anyway.
Push down the price to $1,000 or less, a target at which many scientists and technologists are aiming, and testing for errors in an individual's genetic code, which may be at the root of a variety of illnesses, could become a common diagnostic procedure.
"For sure once and then probably check every few years or every 10 years, depending on your insurance," University of Illinois Professor Aleksei Aksimentiev said recently.
Aksimentiev, UI electrical and computer engineering Professor Gregory Timp and colleagues might have one way of sequencing DNA molecules quickly and inexpensively – by nudging them through pores a billionth of a meter in diameter built into a silicon chip.
The nanopores act as a capacitor that interacts with the natural charge of the molecules when they're moved back and forth through the channels by varying the voltage running through electrodes on either side.
The nanopores are so small, and the DNA molecules so relatively big in comparison, the molecules need a shove or they probably would get stuck in the channels.
The motion of the DNA through the pores depends on its sequence, more or less the relative order of its nucleotides, a chemical, sugar and acid mix that's a basic building block of the molecules. That motion can be captured as a signal – Aksimentiev called it an "electrostatic fingerprint" – and interpreted, theoretically, to reveal a DNA molecule's sequence.
"It is an extremely scalable approach if it works," said Aksimentiev, a physics professor who focuses on computational biology, nanotechnology and biophysics, including easier sequencing of DNA.
One reason: Silicon technology is well-tested and understood. It is, after all, the basis of today's near-ubiquitous microchips, which also happen to be festooned with nanoscale circuitry and other tiny features, the manufacture of which likewise is now commonplace.
"We're using the same kind of technology that's being used by Intel to make state-of-the-art microprocessors," Timp said.
That doesn't mean the effort is without significant challenges. Intel, for example, doesn't make chips to operate in the kind of saline solution environment typical for working with DNA.
"We have to find ways of getting around those problems," Timp said.
For now, the new DNA-sequencing technology has proven itself in complex computer simulations developed by Aksi-mentiev's UI lab. The results were outlined last month in the Journal NanoLetters on the Web by Aksimentiev, Timp, postdoctoral researcher Grigori Sigalov, the paper's lead author, and graduatestudent Jeffrey Comer. The work was funded by the National Institutes of Health and the UI.
But Timp's lab recently fabricated chips with capacitor capability and nanopores, and researchers are now running DNA through the channels for real. The signal they're getting is good, Timp said.
He said the challenge ahead is making sense of those signals as they relate to the sequences of differing strands of DNA.
Aksimentiev said the researchers are working on ways to reduce the "noise" from other activity on the molecular level and get an even clearer signal from the DNA – hence a clearer picture of its sequence.
They're also looking at techniques to slow the movement of the molecules through the nanopores, the better to "see" them because there would be more time to take measurements, perhaps even on a nucleotide by nucleotide basis.
Timp said the researchers use powerful electron microscopes to track the action – plus Aksimentiev's computer simulations, which he likened to a virtual, high-resolution microscope.
Given that the researchers are looking at tiny molecules, it might seem counterintuitive that it takes a lot of computing power to do so. Aksimentiev and colleagues run their simulations both on a supercomputing cluster shared by the UI physics and materials science programs and on the clusters at the UI-based National Center for Supercomputing Applications.
DNA molecules are, in essence, a collection of nucleotides that may vary by just a few atoms. Simulating all those atoms isn't unlike cosmological simulations involving millions of stars, which also are heavy consumers of supercomputing power.
Moreover, the researchers track the movement on a time scale of a femtosecond, that is, a millionth of a billionth of a second. (Put another way, a femtosecond represents the same portion of a second as a second does to 32 million years.)
That kind of frequency can make for a billion steps or more in one of the simulations, Aksimentiev said.
"It's a dynamic system," Timp said. "It's changing all the time."