Programming smart molecules
Computer scientists have put powerful probabilistic reasoning algorithms into the hands of bioengineers.
The Harvard School of Engineering and Applied Sciences (SEAS) and the Wyss Institute for Biologically Inspired Engineering at Harvard University have joined forces to give bioengineers some workng tools of advanced math.
In a new paper presented at the Neural Information Processing Systems conference on December 7, Ryan P. Adams and Nils Napp have shown that an important class of artificial intelligence algorithms could be implemented using chemical reactions.
These algorithms, which use a technique called “message passing inference on factor graphs,” are a mathematical coupling of ideas from graph theory and probability.
They represent the state of the art in machine learning and are already critical components of everyday tools ranging from search engines and fraud detection to error correction in mobile phones.
Adams’ and Napp’s work demonstrates that some aspects of artificial intelligence (AI) could be implemented at microscopic scale using molecules. In the long term, the researchers say, such theoretical developments could open the door for “smart drugs” that can automatically detect, diagnose, and treat a variety of diseases using a cocktail of chemicals that can perform AI-type reasoning.
“We understand a lot about building AI systems that can learn and adapt at macroscopic scales; these algorithms work behind the scenes in many of the devices we interact with every day,” says Adams, an assistant professor of computer science at SEAS whose Intelligent Probabilistic Systems group focuses on machine learning and computational statistics. “This work shows that it is possible to also build intelligent machines at tiny scales, without needing anything that looks like a regular computer.
'This kind of chemical-based AI will be necessary for constructing therapies that sense and adapt to their environment. The hope is to eventually have drugs that can specialize themselves to your personal chemistry and can diagnose or treat a range of pathologies.”
Ryan P. Adams, assistant professor computer science, Harvard School of Engineering and Applied Sciences (SEAS).
Adams and Napp designed a tool that can take probabilistic representations of unknowns (probabilistic graphical models, in the language of machine learning) and compile them into a set of chemical reactions to estimate quantities otherwise unobservable.
The key insight is that the dynamics of chemical reactions can be mapped directly into two computational steps computer scientists normally perform via computer simulation [in silico].
This insight opens up interesting new questions for computer scientists working on statistical machine learning. One being, how to create algorithms and models tackling the uncertainty molecular engineers typically face. In addition, addressing the long-term possibilities of smart therapeutics. Basically, open the door for converting natural biological reaction pathways and regulatory networks into datasets useful for drawing conclusions.
Just like robots, biological cells must estimate external environmental states and act on them; designing artificial systems that perform these tasks could give scientists a better understanding of how problems might be solved on a molecular level inside living systems.
“There is much research on how to develop chemical computational devices,” says Napp, a postdoctoral fellow at the Wyss Institute, working on the Bioinspired Robotics platform, and a member of the Self-organizing Systems Research group at SEAS. Both groups are led by Radhika Nagpal, the Fred Kavli Professor of Computer Science at SEAS and a Wyss core faculty member.
At the Wyss Institute, a portion of Napp’s research involves developing new types of robot devices that move and adapt like living creatures.
Explains Napp: “What makes this project different is that, instead of aiming for general computation, we focused on translating particular algorithms already successful at solving difficult problems in areas like robotics — into molecular descriptions.
“For example, these algorithms allow today’s robots to make complex decisions and reliably use noisy sensors. It is really exciting to think how these tools might be able to help build better molecular machines.”
The field of machine learning is revolutionizing many areas of science and engineering by extracting useful insights from vast amounts of weak and incomplete information. It is not only fueling an interest in “big data,” but is also enabling rapid progress in more traditional disciplines such as computer vision, estimation, and robotics, where data is available but difficult to interpret.
Bioengineers often face similar challenges. Many molecular pathways are still poorly characterized and available data is corrupted by random noise. Using machine learning, these challenges can now be overcome by modeling the dependencies between random variables and accumulate any small amount of information each random event can provide.
“Probabilistic graphical models are efficient tools for computing estimates of unobserved phenomena,” says Adams. “It’s very exciting to find that these tools map so well to the world of cell biology.”