Research on living systems at SFI spans: the origin of metabolism from early-earth geochemistry; the integration of energy capture, reproduction, and mutation in artificial organisms; the creation of minimal forms of life; the core principles governing ecosystem construction, stability, and measurement; the mechanisms providing stability at the social level; and applications of phylogenetic methods to vaccine development for HIV.  Ecological integration is a pervasive theme, from suggestions that the ecology-level metabolic network formed before the emergence of cells, to the recognition that individuality is an emergent concept within ecological frameworks, and culminating in system-level analysis of ecosystems and social systems (ecologies of behavior), which incorporate their network structures and long-range constraints together with individual-based dynamics. Many projects involve generative models and first-principles prediction, and emphasize the emergence of informational elements at the system level that are not reducible to properties of individuals.  Summarized here are four main research areas. For a comprehensive list of SFI Researchers working on topics related to the Dynamics of Living Systems, click here.

• Origin, Synthesis & Form of Life

• The Emerging Ecology of Living Systems

• The Emergence of Social Ecosystems

• HIV Propagation and Treatment

Origin, Synthesis & Form of Life

The Origin of Life Program at SFI seeks to understand how life began and its essential properties. To address these questions, SFI researchers have adopted a three-pronged approach, asking whether life is a natural and perhaps necessary outgrowth of first principles in physics and chemistry, whether life can be synthesized, and what minimal life forms, like viruses, can reveal about life’s fundamental properties.

 SFI Professor Eric Smith and External Professor Harold Morowitz, in collaboration (supported by the NSF FIBR program) with Shelley Copley, argue that far from being an unlikely accident triggered by the first formation of self-replicating RNA, life is likely a natural and perhaps necessary outgrowth of geochemistry, and the forces responsible for its emergence can still be seen in the organization of the major universal core metabolic pathways and structures of cells. Whereas modern life is dependent on highly evolved enzymes for most biochemical reactions, consideration of the substrate molecules (the metabolites acted on by enzymes) and their reactions suggests a self-organizing system, whose structure is tightly linked to the limited chemical energy sources that would have been available on the early earth.  One goal of Smith and Morowitz is to show that in the emergence of life, the major outlines of metabolism were established by chemical kinetics in a world of small molecules, and later refined but largely not changed as proteinaceous enzymes and polynucleotide genes emerged.

Another outstanding problem in research into the origin of life has been to derive a definition of life that satisfies two criteria -- one inclusive and one exclusive. The exclusion criterion should dichotomize the physical universe into living and non-living systems; whereas the inclusion criterion should accommodate all compelling empirical examples of life -- from bacteria through to protists, algae, flies, and mammals. SFI External Professors Steen Rasmussen and Norman Packard and colleagues are attempting to synthesize a simple life form with the following hypothesized critical properties: replication, adaptability, metabolism and cellularity. SFI Professor David Krakauer argues that many aspects of life can be studied through "minimal" forms such as viruses, which possess simple cellularity through the capsid or envelope, are capable of adaptive proliferation in appropriate environments through RNA and DNA replication, and derive energy by parasitizing host metabolism. Viruses occupy a special place in nature, as they threaten to dissipate host catalytic cycles. The virus life cycle raises important questions about levels of selection and individuality, the quantitative cost of reduced autonomy, and the role of noise in evolutionary dynamics. Individuality can be thought about in terms of statistical sufficiency and makes a natural connection to the physical concept of a macrostate. Krakauer and colleagues are seeking to demonstrate that biological individuals behave as macrostates -- partitions of phase space of a minimal statistical complexity predictive of their success into the future.

Return to top

The Emerging Ecology of Living Systems

Any deep account of how individual organisms and species come to be distributed across landscapes must also consider the organization and dynamics of interactions among taxa. Such interactions provide the biotic framework for the flow of energy and resources within and across ecosystems.  Ecological interactions and flows are particularly important for understanding how diversity is generated and maintained, as well as understanding the fate of species and ecosystems in response to natural and anthropogenic perturbations, such as climate change, habitat loss, and invasions.

SFI Visiting Professor Jennifer Dunne, External Professor Mercedes Pascual and colleagues at the Pacific Ecoinformatics and Computational Ecology Lab study the emergence, organization and dynamics of complex networks of interacting species. Whereas quantitative analysis of food-web structure has a three-decade history in ecology, it is only recently, partly through the efforts of SFI researchers, that this work has been drawn into the broader context of network research. This has led to new insights about characteristics, generalities, and scale-dependencies of ecological network structure, and has also identified limits to prior claims about the ubiquity of network patterns such as power-law degree distributions. Dunne in collaboration with SFI Professor Doug Erwin is now extending the study of ecological networks through deep time, as well as their role in niche construction. Initial findings suggest that the fundamental network structure of modern and ancient food webs is very similar, that such structure scales with species and link richness, and that observed structures are well-described by a simple niche model of food webs. Ongoing and future work is focused on compilation and analysis of multiple food webs across the Phanerozoic, including more focused work to look at whether and how network structure responds to major environmental shifts, trophic habit innovations, and major extinctions.

In addition to studying ecological networks, SFI researchers are engaged in modeling ecological processes. SFI Postdoctoral Fellow Michael Gastner is extending the well-known Contact Process model for colonization-extinction processes in space to population dynamic processes. SFI Postdoctoral Fellow Lauren Buckley has been developing dynamic bioenergetic models to predict species’ distributions and range shifts. The models are among the first to produce dynamic range predictions based on first principles of morphology and physiology. Future work by Buckley will extend and empirically validate these models, in particular comparing the ecologies of endotherms and ectotherms, and will include the first analysis of environmental and historical constraints on global patterns of amphibian richness, a widely recognized but poorly understood indicator of the biodiversity implications of environment change. Postdoctoral Fellow Joshua Ladau is developing null models to test for the effects of competitive interaction on community composition. These methods are robust and account well for much of the variability observed in real ecosystems. Ladau is currently extending these methods to treat multiple hierarchical levels of community structure, and developing a public-domain computer application to standardize tests for effects of interspecific interactions in ecological communities. In addition to being useful for testing for the effects of competitive interactions, the methods appear to be one element in a simple, general theory of community assembly. SFI Professor Geoffrey West, External Professor Jim Brown , and SFI Postdoctoral Fellow Chen Hou are integrating ecological principles with results on the thermodynamic organization of metabolism, an extension of very successful SFI research on allometric scaling (see Physics of Complex Systems).

Return to top

The Emergence of Social Ecosystems

Research at SFI on the emergence of social ecosystems has two major components: the role of conflict and the role of communication.

Most communication studies consider the role of signaling in coordinating or manipulating sender-receiver behavior. The information flow is strictly pair-wise, and it can be conceived of as dyadic coding. Also of interest, and a research focus of SFI Professor David Krakauer and SFI Research Fellow Jessica Flack, is the role of communication networks in the emergence of group-level properties that have a delayed and long-lasting influence on individual behavior. This question requires consideration of network coding --how information flows over a network of signaling interactions. Of particular interest is the role of network coding in buffering against errors at the dyadic level and how the capacity of nodes to process increasingly nonlocal information through network coding influences robustness and complexity. Krakauer also works on the evolution of grammars in human language and at the genomic level. Flack additionally studies how receivers generalize from prior associations to decode novel signals when the referents of those signals are spatially or temporally divorced from the signal.

All biological and social systems are comprised of components, or actors, with partially overlapping interests. When component interests are not perfectly aligned or when information is imperfect, conflict inevitably arises. ‘Conflict’ in this sense refers to interactions characterized by an asymmetrical payoff matrix, or those in which individuals rank the set of possible interaction outcomes differently. The role of conflict in facilitating or impeding the emergence of new biological units is of particular interest. Krakauer, Flack , SFI External Professor Nihat Ay, and collaborators have been studying the multi-scale network dynamics of behavioral conflict in relation to how behavioral strategies for managing conflicts evolve in systems in which interactions are polyadic (involving multiple individuals rather than being simply pair-wise) The development of a new, network conflict theory, to include the development of new measures of causality and information flow in networks are in their early stages, and new methods of data analysis of non-linear time series over networks, informed by careful measurements of conflict in a model system (macaque society), are important goals of the project. SFI Professor Jon Wilkins works on closely related issues in the context of intragenomic conflict.

Return to top

HIV Propagation and Treatment

Successful application of mathematical models to the evolution of microbes, and HIV in particular, has long been an important activity at the Santa Fe Institute. Historic HIV samples have rarely been available from archived tissue or blood samples, the oldest positive sample originating from the Congo, and dating from 1959. Therefore, evolutionary models are critical for helping us understand the history of HIV epidemic in the human population, a virus now estimated to have infected 75,000,000 people, but which was only discovered in 1983. In 2000, SFI Professor Tanmoy Bhattacharya and SFI External Professor Bette Korber tracked the origin of the HIV epidemic in humans to the years 1915-1941, providing evidence against the controversial hypothesis that it was introduced into the human population through contaminated polio vaccines. A major limitation to current phylogenetic analysis is that a heuristic search is carried out and the one history found that best explains the data is treated as the true phylogeny.  It is highly desirable at this juncture, when the field is moving away from being data limited to being analysis limited, that we properly estimate and take into account the error resulting from this. Korber and Bhattacharya will implement a Markov Chain Monte Carlo sampling of the Bayesian posterior distribution of the phylogenies and tune the simulation parameters for the available HIV data.  Such a study has not been carried out for this magnitude of data and for the needed evolutionary models before

On the practical side, the models Bhattacharya and Korber have developed provide ideas for vaccine design. The existing diversity of HIV is a serious impediment to developing vaccines against it.  Both T cell and B cell (antibody) responses should optimally be elicited by an HIV vaccine, and they have very different characteristics; T-cells recognize short peptides cleaved from intact viral proteins, while the potent antibodies that can neutralize the virus generally depend on the conformation of the folded protein.

Return to top

Please direct comments on the content of this page to Jessica Flack ().