Skip Navigation

Mathematics and Natural Sciences

Berry Brosi
Environmental Science
The Consequences of Species Extinctions for the Structure of Mutualistic Species Interaction Networks

This proposal is focused on understanding the consequences of species extinctions for the structure of plant-pollinator networks (the connections between plant and pollinator species). We propose to use manipulative field experiments involving the temporary removal of single pollinator species from field plots to assess the impacts of species losses on pollination networks. This work builds from previous work in my lab conducting essentially identical field manipulations, but with different target measurements assessed (previous work focused on foraging behavior and pollination efficiency of remaining bees in the system). We hypothesize that species removals will affect two statistical metrics that describe the structure of polllination networks: first, they will increase network connectance (the proportion of realized links between plant and pollinator species). Second, we hypothesize that the removal manipulations will decrease network nestedness, or the tendency for specialist species to interact with a subset of the partners of generalist species. We expect both of these patterns due to the increased niche breadth (i.e. more plant species visited) of bee individuals that we have previously recorded. This work will be among the first mutualistic (positive ecological interactions) network studies to use an experimental, rather than comparative approach; it will also be rare among empirical network studies in being highly replicated. It will help to form a bridge between network models and empirical data. This work is specifically targeted at collecting pilot data for NSF funding.

Vincent Conticello
Designable Protein Assemblies

Molecular self-assembly is a fundamental principle of life, with cells having mastered this process to encode incredible diversity of function. Helical protein assemblies, including collagen, keratin, F-actin, and tubulin, among others, organize much of the intracellular and extracellular structure, and direct all movement, e.g., flagellate locomotion, cytoplasmic streaming, muscle contraction, etc. The ability to emulate such functions by designing synthetic protein assemblies would transform modern molecular science, with far-reaching applications including locomotion, controlled release, directional transport, dynamic switching, and shape-selective catalysis. However, structurally ordered supramolecular materials on the nanometer length-scale are the most challenging to rationally construct and the most difficult to structurally analyze. The size of these extended protein assemblies and the complexity of inter-subunit interactions present a significant challenge to current computational design methods. In this proposal, we will establish, validate, and make available to the community a novel framework for the targeted design of synthetic protein assemblies at atomic-level accuracy. On the way to developing our framework, we will answer fundamental questions of acute significance to biology and biophysics: from understanding of the functional roles of helical protein assemblies in biology to shedding light on the robustness of protein quaternary structure in sequence space. This proposal will provide a mechanism to promote a collaboration between three investigators with complementary research interests that will result in preliminary data that will support a competitive extramural grant application. 

Alessandro Veneziani
Mathematics and Computer Science
Numerical Methods and Flows at Moderate Reynolds Numbers in Left Ventricular Assisted Devices (LVAD)

Heart Failure has emerged as a major epidemic of this century affecting millions of Americans.  Therapeutic options include left ventricular assist devices (LVAD). this consists of a pump working between the left ventricle and the aorta.  Originally intended to be a bridge-to-transplant therapy, LVAD are nowadays often a destination solution or even a bridge-to-recovery.  Emory University Hospital is an active center of LVAD implantation.  The final outcome of the surgery depends on many factors.  Some of them are related to the dynamics of blood induced by the device.  A possible, non invasive way for investigating these effects in a patient specific setting relies on mathematical models and numerical simulations (Computational Fluid Dynamics - CFD). Morphology and Hemodynamics of the patient can be virtually reconstructed and simulated to assess the performances of the intervention before the real surgery is done. In the specific case, LVAD induces the presence of moderate turbulent flows that require special numerical techniques.  This project concerns the development of numerical methods based on the so-called Large Eddies Simulation (LES) approach for modeling the turbulence. The numerical solver will be used to analyze extensively Emory patients undergoing LVAD for elucidating the complex interplay between the post-surgery morphology and the hemodynamics.  The ultimate goal is to provide surgeons of a mathematical tool to identify the optimal shape that may guarantee the best outcome of the operation.