Pages: pp. 18-19
The juxtaposition of computation and biology opens up a new world of science and technology. Richard Feynman characterizes the young and fast-developing world of computer science as follows: "[it] is like engineering—it is all about getting something to do something." 1 Viewed from this perspective, the scope of research and development at this intersection is a vast, two-way street—what computer science has to offer to biological science and biotechnology and vice versa. Computational thinking helps characterize, predict, and influence the dynamics of biological processes from molecular to cellular to organ in a way that revolutionizes our understanding of health and drug design. In turn, understanding the architecture and principles of natural biological processes and organization might require new models of computation, which could lead to robustness in the design of large-scale software and hardware systems, a hitherto elusive goal.
Today, two key areas drive this convergence between computation and biology. First, the post-genomic challenge, which is the creation of computational models that characterize a natural living cell's inner workings. We now know how to crack an organism's genetic code through sequencing technology. As we begin the 21st century, the next major challenge is to model the genetic program that is executed through gene-protein interactions in a way that characterizes the spatiotemporal dynamics of cellular events. Such models can assist in predicting and controlling responses to external agents and in recognizing highly selective targets and drug design. The task of harnessing the open-source community's power to develop models for intracellular dynamics, cell-to-cell signaling, and even organism organization is a Herculean one, perhaps on a par with the recently completed national human genome project. A DARPA project called BioComputation (see www.darpa.mil/ito/research) is the latest attempt at this grand challenge. A key question to address is the issue of whether the metaphor of circuits and networks is rich enough to deal with the amazing subtleties and complexities of biological systems.
Second, as we reach the limits of Moore's law and look beyond silicon for novel substrates to perform computations, information processing, and storage, biomolecular mechanisms present a potentially revolutionary alternative. For example, we could code information in DNA fragments and carry out complex information processing with nucleotide operations such as ligation, restriction, and hybridization in a potentially massively parallel fashion. Since Len Adleman's seminal work in 1994, which showed the potential of DNA computing for complex problems such as the traveling salesman problems, impressive ideas and developments have emerged such as the solution of six to 10 variable satisfiability problems and tagged DNA storage with thousands of elements. Although building a DNA Pentium chip is still an elusive goal, an area that holds much promise is the design of computationally driven, precisely engineered nanostructures that exploit DNA self-assembly. The essential idea is to produce arbitrary two- and three-dimensional structures from many smaller, information-rich structures that can carry the self-assembly code. Such structures can help us build molecular cages for crystallography, layout molecular electronic devises, and create new materials. As with other new computational substrates such as quantum, spin, or molecular electronics, it is simply too early to tell where this work will lead us. However, there are enough signposts pointing to revolutionary capabilities to give us optimism about the serendipity in discovering new technologies.
This special issue on biocomputation brings you three articles that address challenges, opportunities, and ongoing research work. In the first article, Rajeev Alur, Calin Belta, Vijay Kumar, Max Mintz, George J. Pappas, Harvey Rubin, and Jonathan Schug address the challenges of modeling biomolecular networks and cellular circuitry, capturing the complex nature of interactions including continuous, discrete, and asynchronous dynamics. In the second article, John H. Reif surveys the exciting research area of DNA-based programmable nanostructures, possible applications, and the challenges ahead. Finally, the article by Bud Mishra focuses on the interplay of DNA manipulations and silicon processing and how an effective combination of the two can help solve the genome comparison problem.
We hope this special issue conveys the same sense of excitement that we have felt in the last two years and that it stimulates you to think further about the emerging area of biocomputation.
Thanks to Jordan Feidler of Mitre for his help in preparing this special issue.