Complex Systems in Biology

by Marina Chicurel

(Posted September 18, 1998 · Issue 38)

How many times has a scientist designed a clever and elegant experiment, expecting it to yield a direct and clear-cut answer to a biological question? And how many times has this naive and simpleminded prediction let an experimenter down? Experimental biologists are continuously reminded of the stunning and beautiful, but often seemingly overwhelming, complexity of biological systems. Hardly a day goes by in the lab in which scientists refrain from cursing one or another aspect of the complex nature of biology. Yet its complexity is precisely what makes biology so fascinating.

So how do biologists cope with this love-hate relationship? The development of a powerful set of computational and mathematical tools promises to revolutionize the way scientists study the complexity of biological systems. Researchers' understanding and ability to manipulate and predict many complex behaviors is rapidly increasing thanks to the modeling of simple interactions between numerous biological elements. And because the same tools can be applied to a wide variety of systems, including physical, chemical, biological, social, economic, and even business and political systems, advances in one field often stimulate progress in others. Thus, it seems likely that the study of complexity will advance rapidly, and as new and global insights emerge, the approach to biological research may change profoundly.

The Web is an excellent resource for beginners learning about complex systems. It is also a source of tools, technical information, and a means of communication for complexity experts. Although very few sites are specific to the subject of complexity in biology, many examples of complexity research as applied to biological systems are available at the richly multidisciplinary Web sites of research institutes and centers.

Educational Sites for Beginners

Defining a "complex system" is not an easy task. Participants at last year's International Conference on Complex Systems actually spent several hours debating this definition. Most experts agree, however, a complex system is a dynamic system that behaves in a nonlinear fashion and is composed of many interacting elements. Emergent behaviors - global behaviors that arise but are not easily predicted from the simple interactions of the individual components - are their hallmark. Thus, one can readily categorize many biological systems, including biomolecules, cells, tissues, organisms, and groups of organisms, as complex systems.

The Connected Mathematics: Making Sense of Complex Phenomena Through Building Object-Based Parallel Models project, managed by the Center for Connected Learning and Computer-Based Modeling at Tufts University, is perhaps one of the best educational Web sites on complex systems. One of the main goals of this project is to build tools that will help learners, at all levels, understand complex systems. The project's designers hold that since complexity involves nonlinear and emergent properties, and is a subject that extends beyond the boundaries of traditional disciplines, its explicit form is rarely found in K-16 curricula. Yet a strong need exists, both in the general public and in the scientific community, for the development of skills in systems thinking. The Connected Mathematics project seeks to remedy this situation. A highlight of this site is a library of downloadable "connected models" - interactive computer programs made with object-based parallel modeling languages that simulate the behavior of complex systems, which users can manipulate and extend.

Another educational site is the Cognition,Technology, and Complex Systems Web site. This site, part of a project based at the University of Georgia and funded by the National Science Foundation, is designed to investigate how people can understand and learn the fundamental concepts of complex systems. The site provides a list of books including several didactic books for beginners, and an Overview section that offers a good glossary of frequently used terms. The Exploratorium CompLexicon, which forms part of the 1997 "Turbulent Landscapes" exhibit by the Exploratorium museum of arts and sciences in San Francisco, is another helpful source of basic definitions.

A site with excellent explanations for both basic and advanced concepts in nonlinear science is the Frequently Asked Questions document from the Usenet newsgroup sci.nonlinear. Put together by Jim Meiss at the University of Colorado at Boulder, this document features not only clear and precise explanations in its Basic Theory and Applications and Advanced Theory sections, but also maintains a large and well-categorized list of books and articles that includes popularizations, basic texts, advanced texts, and instructional articles. Its list of Web sites is extensive and is well organized into subtopics.

Many kinds of complex behavior can be simulated by a network of identical interacting elements that obey simple rules. These types of arrays are called "cellular automata," and they constitute a powerful and widely used tool for modeling complex systems. Cellular Automata offers a review of Web sites, including links to basic and advanced tutorials. An example of the use of cellular automata is the modeling of artificial life (or Alife). As explained by the site Alife Online, artificial life is a discipline that attempts to recreate biological behaviors using computers and other "artificial" media. By trying to build systems that behave like living organisms, artificial life complements the traditional approach of taking apart such organisms to see how they work.

Another entertaining and interactive way to gain an intuitive grasp of the power of these tools is to visit the Temple of Alife. Created by the artists at TPR/fusebox, the Temple of Alife features several interactive demonstrations of artificial-life algorithms that simulate a variety of complex behaviors. This author particularly recommends the Planet Wator section, where one can observe the dynamic behavior of a coexistent population of fish and sharks, and manipulate their populations through modifications to the gestation periods of either the fish or the sharks, their size, their initial numbers, or even the size of the sharks' stomachs.

General Resources

Complexity On-line is probably the most comprehensive information service on the Web for all issues regarding complexity. Coordinated by the Charles Sturt University in Australia, Complexity On-line is a network of servers that provides a unified framework for information on complex systems. Of particular interest is the Complex Systems Virtual Library, where one can find information classified by resource type, research topic, or country. Other valuable resources within this site include a list of newsgroups, a searchable bibliographic database, and a software repository. The announcement and call for papers for a conference entitled Complexity Between the Ecos: From Ecology to Economics, are also available at this site. The conference is to be held at the University of New South Wales in Sydney, Australia from November 30 to December 4, 1998. Finally, Complexity International, a refereed journal published by the Johnstone Centre at the Charles Sturt University, is also worth exploring.

The Cognition,Technology, and Complex Systems Web site mentioned above is also a highly up-to-date source of general information on complex systems. Its lists of references, software applications, and Web links are of particular interest. Though less extensive, the Yahoo Index of Complex Systems Sites is another good source of Web links. Finally, the Institute for Information Technology in Canada maintains an extensive subject index on artificial intelligence Web links, including Agents, Artificial Intelligence (AI), Chaos, Complex Systems, Nonlinear Systems, Education and AI, Evolutionary Algorithms, Genetic Algorithms, Genetic Programming, Fuzzy Logic, General AI, History of AI, Medical Applications of AI, and Neural Networks.

Research Institutes and Centers

The Santa Fe Institute, a multidisciplinary research and education center in New Mexico, constitutes one of the hubs of research on complex systems. Scientists from universities and research institutions throughout the world converge at the Santa Fe Institute to pursue research on complexity, with projects ranging from the patterns of communication between ants to the spread of information across economic markets. The Institute is also the home of Stuart Kauffman, one of the leading thinkers on complexity especially as applied to biological systems. A brilliant and very approachable exposition of Kauffman's ideas on complexity can be found in his recent book At Home in the Universe: The Search for the Laws of Self-Organization and Complexity.

Two sites maintained by the Santa Fe Institute are particularly worth visiting: the Evolving Cellular Automata Project (EvCA Project) and the aforementioned Alife Online. The EvCA Project's goal is to understand how evolution produces sophisticated emergent properties in systems composed of simple elements that are limited to local interactions. Alife Online includes FAQs, general descriptions of many topics in Alife, outstanding lists of Web links and software, search capabilities, and information on upcoming conferences, courses, and publications. As of June 30, 1998, this site is being renovated, and promises to improve even further. Not only is it extremely up-to-date, but it now gives visitors an option to receive automatic email notifications on new updates as they become available.

The New England Complex Systems Institute (NECSI) is another organization specializing in the investigation of complex systems. NECSI coordinates and supports eight working groups: information mechanics; time series analysis and prediction, molecular self-organization; mind, brain and behavior; informatics; evolution; structure of human organizations; and biomedical complexity. Each group holds a series of workshops designed to communicate and discuss recent progress and unsolved questions. NECSI also organizes the International Conference on Complex Systems, to be held this October 25-30, 1998, in Nashua, New Hampshire.

NECSI also produces the refereed Web journal InterJournal. The journal differs from many conventional journals in that it makes innovative use of its electronic format to improve the communication of data and ideas between scientists. For example, it accepts a wide range of data formats for publication, including computer programs, raw data, video, audio, and documents structured by hyperlinks. Although the journal specializes in three main subjects - Complex Systems, Genetics, and Polymers and Complex fluids - the vast majority of articles fall into the Complex Systems category.

Two other organizations devoted to research in complex systems are the Center for Complex Systems Research (CCSR) at the Beckman Institute, and the Center for Nonlinear Studies at the Los Alamos National Laboratory. At CCSR, an interdisciplinary group of researchers are investigating complex dynamic processes in biology, physics, chemistry and astronomy. Projects in biology include investigations on a peroxidase-oxidase chemical oscillator, metabolic networks and neural network dynamics. The Center for Nonlinear Studies is also composed of a highly interdisciplinary research group. Two main areas of biological research are "Landscapes and Dynamics in Proteins," and "Probabilistic and Combinatorial Analysis of Biological Systems." The "Biology (Proteins)" link on the home page is under construction, but one can obtain information on both of these projects from the Mission Statement page.

So, the next time one is immersed in a nightmarish effort to understand why the inhibitor of the activator of the regulator of the kinase pathway that converges, diverges, and overlaps with several of one's favorite receptor-activated pathways is turned off, rather than on as predicted, one can get on the Web and learn more about complex systems. It may help more than one might think.

Marina Chicurel is a postdoctoral fellow at Harvard Medical School/Children's Hospital and at the Albert Einstein College of Medicine in New York.

Send us your comments and ideas for future articles.

Endlinks

Evolutionary Programming Society - formed in 1991 to promote research in evolutionary computation and self-organizing systems. The society sponsors the Annual Conference on Evolutionary Programming and maintains the ep-list Internet email digest.

Genetic Algorithms Archive - a repository for information on genetic algorithms (GA), computational simulations of biological evolution that include selection, mutation, and crossovers. This site makes available past issues of the GA-List digest, source code for many GA implementations, announcements about GA-related conferences, and links to other GA-related Web sites. Maintained by Alan C. Schultz at the Navy Center for Applied Research in Artificial Intelligence.

Artificial Life Bibliography of On-line Publications - a recently updated and well-organized site containing over 400 publications sorted by subject. Maintained by Ezequiel A. Di Paolo at the University of Sussex in England.

AI Software Repository - a very extensive collection of software maintained by the School of Computer Science at Carnegie Mellon University. , it was Not updated since February 1995, unfortunately.

The DDLAB or Discrete Dynamics Lab - excellent software applications for the analysis of discrete logical networks. Maintained by Andy Wuensche at the Santa Fe Institute.

Guide to Available Mathematical Software - although not specialized for complex systems problems, this very extensive guide is useful as a general source of mathematical software. Maintained by the National Institute of Standards and Technology, the guide serves as both a cross-index and repository for mathematical and statistical software. It provides centralized access to abstracts, documentation, and source code for software, as well as direct access to multiple Web repositories.

@BRINT - although focused on the application of complex systems to the management of organizations, this site offers many general resources on complex systems including lists of publications, Web sites, and tools.

Molecular Physiology of CNS Development - despite its rather misleading title, this site has useful information on genetic networks. When you arrive, scroll down to the site index to begin navigation of this somewhat confusing site. Of particular interest are Recent Publications, Genetic Networks, and Images: Visualizing Genetic Networks. Managed by Roland Somogyi at NIH.

Breathing New Life into Alife Online - This issue's HMS Beagle Site Review discusses this artificial-life site.

Web sites mentioned in this column:

Educational Sites for Beginners