Matthew Lakin, research assistant professor in Chemical and Biological Engineering and in Computer Science explains the idea of molecular computing.
“You have a processor in your computer that does computation," Lakin said. "We want to design collections of molecules that will do similar computational tasks by reacting with each other based on their chemistry. The particular type of computation that we are interested in is adaptation and learning. We want a system that can learn to adjust its behavior based on things that it sees floating around inside cells.”
Wait a minute. You can train dogs to go find things and report back. Bomb sniffing dogs or drug sniffing dogs do that routinely – but Lakin is talking about training molecules or collections of molecules.
Size is what drives the interest in molecular computing. Molecules are so tiny they can move into and out of cells.
The technical capability to program collections of molecules would allow researchers to build a device that could go into a cell and sense whether the cell is healthy or not. The collection of molecules could then communicate the information back to doctors or it could try to treat the cell directly. Because molecules can interface directly with cells, the researchers are working with what nature has already developed.
“You could imagine a system of nanoscale sensing networks that can learn from their experience," said Lakin. “Based on what they observe, they adapt so they are able to make more sensible decisions in the future. For example, to detect diseases that are constantly mutating.”
If the collection of trained molecules can make a decision that could turn into an action, the molecules could decide that the cell has gone wrong and tell the cell to self-destruct. Alternatively, a decision that more of a useful protein is needed that could help the cell improve the health of the body that contains it.
A research group, combining faculty and students from the departments of Computer Science and Chemical and Biological Engineering is collaborating on new research grants from the National Science Foundation that will enable them to build such molecular systems.
UNM, like most large research institutions, encourages groups of researchers to work together to form questions and find answers. Those groups are frequently interdisciplinary and involve undergraduates as well as graduate students, post-docs and faculty members.
“We’re computer scientists so we write programs and simulations and then we try to integrate that with the experimental people,” said Lakin. “We suggest to them what type of experiment they might want to run to test our ideas.”
The Computer Science researchers use their software tools to optimize the tasks the molecules need to perform and design molecules to perform those tasks. The researchers in Chemical and Biological Engineering are working on experimental implementations of molecular learning circuits in the laboratory. Undergraduate students are participating in this research by performing some of the experiments to test theories, and in some cases helping design the experiments.
Understanding the cellular environment
Chair of the Department of Computer Science and Professor Darko Stefanovic said part of their work involves collaborators at Columbia University and the Hospital for Special Surgery, which is affiliated with Cornell University and at Portland State University. They are trying to understand basic things about how DNA reactions work.
“One of our goals is to understand how DNAzymes work, said Stefanovic. “DNAzymes are DNA strands that have the ability to catalyze the modifications of other nucleic acid strands. In our previous work we have used DNAzymes that are able to assume some three dimensional structure and wrap about another nucleic acid strand such as an RNA strand and help cleave it into two.”
It might be useful to control that action. “One of the common therapeutic tools is called a knockdown, which is to prevent a gene from being activated,” said Lakin. “This approach to cleaving nucleic acid strands and preventing them from producing a protein that might potentially cause something bad to happen in your body is a powerful approach for a disease targeting applications. So it is important to optimize the conditions under which this might take place.”
“We would like to ultimately use molecular computing to sense the kind of things that a cell might sense," Stefanovic said. "What’s in the inside? What is the current state of the cell and the molecules inside?”
There are other possible applications for the research. Stefanovic said it may be possible to make water monitoring technology more efficient by speeding the rate at which potential problems are identified.
Work under the new grants has just begun. Now the faculty members and students are plunging into the detail that makes this research such a challenge. Students who want to become involved in research that can change our approach to human disease or environmental monitoring can begin by looking at courses in Computer Science and Chemical and Biological Engineering. Or they can contact professors Steven Graves, Matthew Lakin and Darko Stefanovic at the Center for Biomedical Engineering.