Emmanuel Tannenbaum, Assistant Professor
Ph.D. Chemical Physics, Harvard University, 2002
Phone: 404-385-0287
Fax: 404-894-0519
Office: Cherry Emerson 313
Research Interests
We are a theory group in the School of Biology at Georgia Tech, with a primary interest in evolutionary dynamics. Briefly, evolutionary dynamics is a subfield of mathematical biology that deals with developing mathematical models describing evolutionary processes in biological systems. In principle, these evolutionary processes can occur at size scales ranging from molecular to organismal, and time scales corresponding to both microevolutionary and macroevolutionary processes.
Because evolutionary dynamics has connections to fields such as game theory, mathematical economics, information theory, and computational neuroscience, in the future we are potentially interested in moving into one or more of these areas as well.
Research in the group is in three general areas: (1) Quasispecies theory; (2) Time and energy costs in biological systems; (3) ``Animate" behaviors in biochemical networks
A description of each of these research tracks is provided below:
QUASISPECIES THEORY
Quasispecies theory refers to a set of ordinary differential equations developed by Manfred Eigen and Peter Schuster in order to model molecular evolutionary processes relevant to the origin of life.
Quasispecies theory has been used extensively to model viral evolution, and more recently has been applied to single-celled life as well. In this context, quasispecies theory could be useful for modeling the emergence of drug resistance in bacteria, mutation-propagation in stem cells and the emergence of cancer, tissue aging, and viral-immune system interactions.
TIME AND ENERGY COSTS IN BIOLOGICAL SYSTEMS
Biological systems can exhibit a wide variety of structures and behaviors that have presumably evolved because they provide a selective advantage to the organism. While some structures and behaviors are highly specialized (the so-called ``poison dart'' frogs for instance), other structures and behaviors are common in organisms at certain levels of complexity.
For example, complex multicellular organisms can exhibit structures and engage in behaviors that simpler, few-celled organisms do not. Examples include sexual replication, sleep, and a stem-cell-based tissue architecture.
As a result, we seek to elucidate both the selective advantages for various common structures and behaviors, and the time and energy costs associated with them, in order to outline regimes where these structures and behaviors are advantageous or disadvantageous. Such studies, if successful, could provide important insights into the various evolutionary pressures driving the emergence of complex terrestrial life.
``ANIMATE" BEHAVIORS IN BIOCHEMICAL NETWORKS
There is emerging evidence to suggest that much of the so-called ``junk" DNA in complex eukaryotic organisms is not ``junk" at all, but rather codes for a vast, RNA-based, genetic regulatory network. Elucidating the structure of this network will be key to understanding the evolutionary processes that gave rise to complex terrestrial life.
In a recent speculative paper, we argued that there should be ``scale-free'' features associated with systems constructed by the self-organization of agents (molecules, cells, people) acting under one or more selection pressures. Therefore, by studying structures and behaviors in one such agent-built system, it may be possible to infer structures and behaviors in another such system.
Since cellular life presumably arose from autocatalytic polynucleotide and polypeptide networks, then presumably it may be possible to infer certain structural motifs in these networks by studying other complex systems, such as the brain and free-market economies. It is believed that brain structure is driven by the self-organization of neurons acting under a chemically-based reward-punishment system, while free market economies are driven by competition for money.
If this ``scale-free'' analogy proves to be correct, then it could become a highly useful approach for elucidating various structural and behavioral motifs in highly complex biochemical networks.
Current Research
Within the first research track, dealing with quasispecies theory, my group is interested in the following problems:
1. Incorporating genetic repair processes into quasispecies models, in order to develop quantitative models for mutation-propagation in cellular organisms. Examples of such repair processes are the SOS response and Mismatch Repair. We plan to also consider mechanisms such as apoptosis, which do not protect an individual cell from genetic damage, but nevertheless prevent the accumulation of mutations in a population.
2. Extending previous models describing the evolutionary dynamics of a population of adult stem cells, to include the effects of multiple chromosomes, and the interaction between the stem and tissue cell population. We plan to consider both random chromosome segregation mechanisms, and the so-called "immortal strand" co-segregation mechanism, which is believed to be the chromosome segregation mechanism in adult stem cells.
These projects, because they deal with mutation-propagation in cellular organisms and in tissues, are applicable to antibiotic drug resistance in bacteria, the emergence of cancer, and tissue aging.
Within the second research track, dealing with time and energy costs in biological systems, we are interested in the following problems:
1. Developing mathematical models describing competition between various asexual and sexual replication strategies, in order to determine regimes where different strategies are advantageous or disadvantageous. This includes gamete differentiation, sex differentiation, the emergence of mating cycles, and various types of mating strategies.
2. Modeling various types of undifferentiated and differentiated labor strategies, in order to elucidate possible selective advantages for structures and behaviors in complex terrestrial life. For example, we hypothesize that the evolutionary basis for a stem-cell based tissue architecture in a complex organism is that it is more efficient, in terms of resource utilization and organismal reproduction rate, if a subset of the cells focus on actually regenerating the tissues, while the remainder of the cells focus only on the tasks associated with proper tissue function (e.g., a skin cell should ``worry'' only about being a skin cell, while regenerating the tissue should be left to a small fraction of adult skin stem cells).
Another example is the phenomenon of sleep. We argue that sleep itself, and the distinct REM and non-REM sleep states, are examples of temporally differentiated labor strategies, whereby a given set of agents oscillate between performing various subtasks associated with a given task. We are exploring the idea that, under certain conditions, temporal differentiation can lead to increased system productivity. In the context of sleep, this implies that sleep emerges because it makes sense for the brain to focus exclusively on one set of tasks related to wakefulness, presumably dealing with the acquisition of information from the environment, and then to switch to another set of tasks, involving the processing and long-term storage of this information. The idea is that the brain can ultimately process more information in this way, which for a big-brained organism possibly leads to a survival advantage.
Within the third, and final research track, we are interested in determining the role, if any, that associative processes play in biochemical networks and genomic evolution. Currently, we are working on models that describe a biochemical implementation of a process known as associative learning, which is believed to be a powerful learning mechanism at work in the brain. Associative learning is essentially a form of learning by analogy. It is a powerful learning mechanism because, by establishing a connection between two seemingly unrelated areas, it allows techniques useful for solving one set of problems to be useful for another set of problems.
One of the most famous examples of this process at work in the brain are the series of experiments by Pavlov, whereby a dog, induced to salivate by being presented with food and the sound of a ringing bell, was induced to salivate by the sound of the ringing bell alone (Pavlov used the term ``conditioning'' to describe this associative learning process).
In addition to developing mathematical models for biochemical implementations of associative learning, we are interested in constructing such systems in vitro (in collaboration with Gonen Ashkenasy, at Ben-Gurion University), and also looking for examples of such processes leading to genomic evolution. Along this line, we wish to explore the possibility that polycistronic RNA in prokaryotes may have arisen via biochemical processes analogous to associative learning or conditioning in more complex systems.
Selected Publications
1. E. Tannenbaum, "Selective advantage for sexual reproduction," Physical Review E 73, 061925 (2006). (PDF)
2. E. Tannenbaum, J.L. Sherley, and E.I. Shakhnovich, "Semiconservative quasispecies equations for polysomic genomes: The haploid case,"
The Journal of Theoretical Biology 241, 791 (2006). (PDF)
3. E. Tannenbaum, J.L. Sherley, and E.I. Shakhnovich, "Evolutionary dynamics of adult stem cells: Comparison of random and immortal-strand segregation mechanisms," Physical Review E 71, 041914 (2005). (PDF)
4. E. Tannenbaum and E.I. Shakhnovich, "Solution of the Quasispecies Model for an Arbitrary Gene Network," Physical Review E 70, 021903 (2004). (PDF)
5. E. Tannenbaum, E.J. Deeds, and E.I. Shakhnovich, "Equilibrium Distribution of Mutators in the Single Fitness Peak Model," Physical Letters 91, 138105 (2003). (PDF)