Seminars

Laboratories certified by the federal Clinical Laboratory Improvements Act to perform laboratory tests on human patients face a number of regulatory, software, hardware and resource challenges with regard to management and utilization of test results for research. The goal of this presentation is to orient the listener to these challenges and provide a framework for discussion surrounding new and innovative ways to overcome obstacles to the utilization of clinical test results for research.

In recent years, cell tracking has become an established research problem, and a number of solutions have emerged that are now being used by biologists. A challenge that is very closely related to the problem of cell tracking is the automatic monitoring of certain time-dependent cellular functions. Fluorescent markers enable the imaging of changes in cell morphology and the monitoring of complex cellular functions such as cell cycle and changes during mitosis like chromosome separation. Our current research focuses on automatic measurements that are based on these dynamic fluorescent markers. In addition, we are investigating the formulation of a dynamic highcontent signature that captures biologically relevant information at the single-cell level. Specific examples demonstrate our ability to characterize cell cycle state. We will outline how this method can be extended to characterize cell death. In addition I will review some ongoing work on the segmentation of densely packed cell populations in 3D confocal image data which is particularly relevant to cancer research. Here model based algorithms are being applied for segmentation that will ultimately lead to the understanding of tissue organization and structure.

Negative feedback is common in all types of biological processes and can increase a system's stability to internal and external perturbations. But at the molecular level, control loops always rely on finite rates for random births and deaths of individual signal molecules. I will show how seemingly mild constraints on the birth rates -- or short delays -- place severe limits on homeostasis and molecular fluctuations that no type of control system could overcome in any context. I also discuss how the limits become dramatically more restrictive in chemical cascades, where information is inevitably lost at each step. The results are formulated in terms of biological observables that were measured for many systems, and I will also discuss some experimental results for bacterial plasmids, based on methods for counting the exact number of molecules per individual cell. The first part of the talk will provide an introduction to random processes in cells.

We have designed a set of protocols that use peer-to-peer techniques to efficiently implement a
distributed and decentralized desktop grid. Incoming jobs with different resource requirements are
matched to system nodes through proximity in an N-dimensional Content-Addressable Network,
where each resource type is represented as a distinct dimension.

We balance load induced by job executions through randomly generated virtual dimension values,
which act to disaggregate clusters of nodes, and also by a job pushing mechanism based on an
approximate global view of the system. We improve upon initial job assignments by using a jobstealing mechanism to overcome load imbalance caused by heterogeneity of nodes/jobs and stale
load information.

We also provide a set of optimizations that combine to minimizesystem overheads created by the
job-monitoring infrastructure, and themanagement of the underlying peer-to-peer system.

An increasing volume of time-stamped patient data in clinical data repositories makes possible the identification and analysis of populations with therapeutic responses, outcomes and clinical care processes that are reflected by multivariate trends and patterns over time. Recognition of these patterns could be quite valuable in clinical and translational research and quality assurance, but is not well-supported by current clinical data retrieval systems. This presentation will propose a temporal abstraction-based query system that supports population data retrieval from clinical databases. An implementation and evaluation of the system's temporal pattern specification and detection component, PROTEMPA, will be described. Possible extensions allowing temporal data mining and analysis will be discussed.

Graph theoretic problems are representative of fundamental kernels in traditional and emerging computational sciences such as chemistry, biology, and medicine, as well as applications in national
security. Yet they pose serious challenges for parallel machines due to non-contiguous, concurrent accesses to global data structures with low degrees of locality. Few parallel graph algorithms outperform their best sequential implementation due to long memory latencies and high synchronization costs. In this talk, we consider several graph theoretic kernels for connectivity and centrality and discuss how the features of petascale architectures will affect algorithm development,
ease of programming, performance, and scalability. Our large-scale graph algorithms are applied to real-world problems in phylogenetic reconstruction of evolutionary histories, inference of gene function in protein interaction networks, and cancer research.

The relatively new discipline of Biomedical Informatics encompasses the application of computer science from the molecular level to the entire US health care system. As our knowledge of life and disease has grown we’ve begun to move rapidly from a descriptive to a quantitative science in which mathematical models running on supercomputers may someday soon provide a powerful model of human life itself. At the same time as we’re making rapid scientific progress, our health delivery system struggles with high cost and poor macro results in large part due to the enormous technical, economic and policy issues associated with achieving ubiquitous, interoperable electronic medical records. In between computing is being applied to medical imaging, to delivering ubiquitous care and to biosurveillance in defense of the public health. This talk provides a broad overview of the field and provides selected examples of some of the most exciting potential opportunities to apply computer science to life, health and disease.

Cancer remains the leading cause of disease death for middle aged Americans. The development of prognostic tools could have immediate impact on the lives of millions of cancer patients. We have developed an integrated, cell-based modeling framework that includes a cellular model for cell dynamics (cell growth, division, death, migration and adhesion), an intracellular regulatory network for cell cycle control, and a continuous equation system for extracellular chemical dynamics.
This model has produced avascular tumor growth dynamics that agree with tumor spheroid experiments and generated experimetally testable hypotheses about tumor microenvironment. In particular, we investigate the mechanisms for tumor growth saturation. Given the biological
realism and flexibility of the model, we believe that it can potentially facilitate a deeper understanding of the cellular and molecular interactions associated with cancer progression and
treatment.

Single-molecule spectroscopy promises the detection and identification of rare events, the measurement of the entire distribution of molecular properties, and the direct study of dynamics. Some recent experimental and theoretical advances that further develop single-molecule spectroscopy towards reaching its full potential will be discussed. Specific examples will be used to illustrate the motivation behind the developments and the new ideas that emerge from such quantitative high-resolution studies.

In anticipation of the 200th anniversary of Darwin's birth and the 150th anniversary of the publication On the Origins of Species in 2009, and in collaboration with the Center for Science Education, the Computational and Life Sciences Stragetic Initiative, and the Origins Chemical Bonding Center at Emory University, we are pleased to announce this public symposium, Evolution Revolution: Science Changing Life.

Dr. Wilson will give his keynote address, Darwin and the Future of Biology, October 23, 2008, 7:00 p.m., Glenn Auditorium.