Engineered nanopores can be employed in a variety of biotechnology applications, including molecular sensing. They also make good DNA and RNA sequencers. Such techniques involve electrically detecting a DNA or RNA molecule as it translocates through the pore by measuring either the voltage across the membrane or the current through the pore.

Measuring ionic blockade signatures

In this study, researchers led by Meni Wanunu in the Department of Physics at Northeastern University and Stuart Lindsay in the Department of Physics at Arizona State University measured the ionic current through a nanopore using a high-bandwidth amplifier as a molecule of tRNA passed through it.

“When a voltage bias is applied across the nanopore membrane, an ionic current is generated through the pore,” explains Wanunu. “This voltage bias electrophoretically drives the negatively charged tRNA though the pore. Since tRNA molecules are bent they must deform as they traverse the pore, a process that takes as long as milliseconds. We measured the so-called ionic blockade signatures in detail during each tRNA molecule’s passage though the pore and repeated these measurements for thousands of molecules.”

Machine-learning algorithm identifies individual molecules

The researchers also employed a machine-learning algorithm (the same one that powers IBM’s Watson) to identify individual tRNA molecules in a sample. “This algorithm allows us to separate the relative contributions of various signal features in the sample and so assign an identity to single tRNA molecules,” explains Lindsay. “The technique also allows us to make use of the time-dependence of signal features to calibrate changes in the nanopore over time, and so correct for them,” he says.

“We use control data from samples of known tRNAs to teach a computer program the unique characteristics that each tRNA has as it translates through a specific type of nanopore,” adds lead author of the study Robert Henley, also at Northeastern. “These characteristics include different features of the noise spectrum that each tRNA produces, and the amount of time it takes to deform from its original conformation and translocate the pore. Once the computer program has learned about these parameters, it is able to optimally identify molecules in an unknown mixture.”

Technique could be used to directly sequence RNA

The researchers say that their work could help in the rapid characterization of tRNA levels in RNA cell extracts, especially in cases where only a small amount of tRNA is available, or when it is necessary to differentiate between tRNA “isoacceptors” or “isodecoders”. “One such example of when this could be particularly interesting is the rapid quantification of the tRNA-ome of different biological cells, and how such global tRNA patterns respond to outside perturbations, such as viral infection or oxidative stress,” explains collaborator Barry Cooperman.

“Given the small, finite number of tRNAs in any sample, our technique could be used to directly sequence RNA,” Wanunu tells

The research is detailed in Nano Letters DOI: 10.1021/acs.nanolett.5b03331.

For more on DNA sequencing, visit the Nanotechnology focus collection at