Networking the “Invisible Colleges”: Application
of Network Theory to Biocomplexity
An National Science Foundation-Sponsored
Biocomplexity Incubation Workshop
Hosted by:
Jeff Johnson, Joe Luczkovich,
Bob Christian, East Carolina University
and
Steve Borgatti, Boston College
To be held
21-24 March 2001, Duke University
Marine Lab, Beaufort, NC
Introduction
There are many disciplines engaged in attempts
to understand and model complex systems of one kind or another (e.g., biological
systems, human systems, neurological systems). This workshop seeks
to solve some of the current problems in biocomplexity by bringing together
a collection of mathematicians, biological scientists, and social scientists
for the purpose of working on modeling complexity in both human and biological
systems with a particular concern for system integration. There have been
researchers in both the biological and social sciences who have been working
on a number of related problems in the area of structural complexity. For
the past 30 years a whole approach in the social sciences, known as social
network analysis, has evolved that focuses on structural models concerning
human interactions. Similarly, a number of ecologists have been working
on various aspects on the network structure and dynamics of energy exchanges
(e.g., carbon). In addition, others in the area of food web analysis
have suggested and applied graph theoretic approaches to the study of trophic
interactions. Despite the similarities in problem area (i.e., the modeling
of complex structures) there has been little communication between the
various groups. A recent exception has brought social and biological scientists
together that has promise for improving the relevance of ecological concepts
through the application of mathematical models developed in the social
sciences. This has led to further interactions between social scientists,
protein chemists, and molecular biologists in developing generalized visual
and exploratory tools for the study of complex systems. The foodweb diagram
below (from Johnson et al. in review) shows the results of one such visual
model using the data from the food web of a tropical rainforest (Reagan
and Waide 1996). This increased communication among these groups
is exciting and holds promise for the acceleration of solutions to various
problems in the study of biocomplexity.
Goals and Objectives of the Workshop
The overall goal of this workshop is to bring
together scientists of various backgrounds to address a number of important
issues in the study of biocomplexity. There are five specific objectives
that relate to specific areas of interest. The specific objectives and
how each will be addressed are given below:
1. Establishing common ground: To
identify the common issues of complexity across disciplines, the limitations
to cross fertilizations of ideas, and ways to overcome these limitations.
-
This will be addressed by providing appropriate
readings prior to the workshop. All participants will receive review articles
from different disciplines and a statement of purpose, prospectus and lexicon
written by the coordinators.
-
The first session will address this topic to
establish commonality among participants.
2. Modeling: To explore the appropriateness
of various mathematical network models and analyses developed by mathematicians
and social and biological scientists for the study of complex biological
systems.
-
Experts from the different disciplines will
present summaries of models developed used in their own disciplines.
Emphasis will be on the application of each approach. Approaches include
input-output analysis, cycling analysis, trophic analysis, information
analysis, and cascade models.
-
Another line of discussion will be on the limitations
currently understood for the models with the expectation that some of these
limitations may be addressed by others through their alternate perspectives.
3. Visualization and Exploration: To
develop tools for the exploratory analysis of complex biological systems.
-
Emphasis of this session will be on accessibility
to software for the models, analyses and visualization of results.
4. Data Issues: To address issues concerning
the quality of data used in modeling bio-complexity in food webs and trophic
networks.
-
Translation of empirical information into models
will be addressed. Issues include the reliability and variability of data,
aggregation of compartments and interactions.
Workshop Format
The workshop will take place during 21-24 March
2001 at the Duke Marine Laboratory, Beaufort, NC. It will be a three-day
affair with each objective being addressed through presentations, breakout
groups on topics covered, final group reports and summaries by rapporteur.