Steve Andrews
Research Interests in theoretical and experimental systems
biology:
My research is in the interdisciplinary field of systems
biology. This field combines physics, chemistry, and biology
methods
to investigate organization within biological systems, on size scales
that typically range from a few proteins to many cells. Results
are yielding deep insights into how the highly structured macroscopic
world of living organisms is built from the stochastic microscopic
world of individual molecules. They are also providing an
improved conceptual foundation for medical and biotechnology
developments, with impacts on topics such as drug discovery,
personalized medicine, biofuel generation, and bioremediation. My
research projects include:
Algorithm development for
cell modeling
Computer simulations are are used in systems biology as a way to build
intuition about the system dynamics, to test hypotheses, to make
predictions, and to
identify essential system components. I am developing
modeling tools that can simulate biochemical systems with a relatively
high level of detail, in which individual molecules are represented
with nanometer-scale spatial
resolution, but that are also fast enough to allow the
simulation of
tens of thousands of molecules over several minutes of real time.
I developed algorithms for simulating
reactions between freely diffusing molecules in solution and for
interactions between
molecules and surfaces. These algorithms are implemented in the
Smoldyn
computer
program, which can be downloaded from the Software page. My
current algorithm development addresses the detailed simulation of
cytoskeletal filaments and membranes.
Macromolecular crowding
Biological cells are highly crowded spaces, with often 20-30% of the
volume occupied by macromolecules such as proteins and nucleic
acids. This fact has been both known and investigated for many
years,
but there remains no predictive theory for how much crowding slows
diffusion, nor for the quantitative effects of crowding on biochemical
reaction rates. Using a combination of analytical theory and
computer simulation, I am developing semi-emipirical theories to
address these questions. If successful, these theories will allow
in vitro measurements of
biochemical reaction rates to be
converted to in vivo reaction
rates, for the appropriate native biological
systems. This will help address a major problem of cell biology
modeling, applicable to both conceptual and computational models, which
is that the quantitative data on intracellular biochemical reaction
rates is generally very sparse.
Cell signaling in yeast
The yeast mating pheromone response pathway is a classic model system
for
studying intracellular signaling because it is relatively easy to
study, is similar to many mammalian signaling pathways, and is a rich
system. In collaboration with other scientists at the Molecular
Sciences Institute, I am using analytical and computational methods to
investigate information transfer along the
pathway. This study addresses topics such as the effects of
biochemical feedback and feedforward on information transfer and the
importance of having aligned dose-response curves at different points
of the signaling network.
Mechanics and dynamics of
the bacterial cytoskeleton
The bacterial cytoskeleton is highly dynamic. For example, the
MinC, MinD, and MinE proteins of E.
coli exhibit a remarkable oscillation between the cell poles:
the MinD protein polymerizes in a helical coil that extends from one
pole towards the cell center, is depolymerized by MinE, forms a new
polymer from the opposite pole, and so on. In another example,
the FtsZ protein forms a central ring around the cell center that
constricts during cell division to yield two daughter cells. In
recent work, I investigated how the micron-scale shapes of these
cytoskeletal polymers might arise from mechanical forces between
individual proteins, along with how these polymers are likely to apply
forces to cell walls. While I am not continuing to focus on this
research direction at the moment, I plan to return to it in a year or
two, equipped with simulation methods that can simultaneously model
chemical reactions, filaments, and membrane dynamics. This cell
division system has been extensively modeled in the past, but it
continues to reveal new insights and to be an ideal model system for
exploring biochemical spatio-temporal dynamics.
© Steven Andrews, 2008. All rights reserved.
Last modified 8/25/08.
