November 2009 |
Shayn M. Peirce, Ph.D.
Assistant Professor of Biomedical Engineering
University of Virginia
"Combining experiments with agent-based modeling to study microvascular growth at the multi-cell level"
Date: Thursday, November 12
Time: 3:00 pm EST
Location: CSEB 320
Abstract:
Understanding where new endothelial cells and perivascular cells come from and which signals mediate their recruitment is central to therapeutically manipulating microvascular growth and remodeling for tissue engineering and regenerative medicine applications. Evidence suggests that vascular cells arise from bone marrow-derived progenitor cells, from within the microvasculature, and from the surrounding interstitial tissue, but it is unclear how these different cell populations contribute to microvascular growth and what signals govern their spatial and temporal dynamics. Cell trafficking—cells moving from one tissue compartment to another via the circulation—is integral to emerging treatments, such as stem cell-based therapeutic angiogenesis, yet cell trafficking is poorly understood in the in vivo setting because technical limitations prohibit the visualization of multiple individual cells in the body and their molecular mediators. We combine in vivo and in vitro experimental assays with agent-based computational modeling to obtain a systems-level view of this complex process and to establish more informed approaches for utilizing progenitor cells to engineer new microvascular networks.
September 2009 |
Dr. Carl Kesselman
Professor, Epstein Department of Industrial and Systems Engineering
Viterbi School of Engineering
University of Southern California
"The Grid as Infrastructure for Sharing Health Information"
Date: Wednesday, September 16
Time: 3:00 pm EST
Location: CSEB 320 with simultaneous webcast (click below to view recorded seminar)
Abstract:
Increasingly, translational research and clinical practice is impeded by the ability to exchange diverse health related information between collaborating parties. The issue of data sharing in a health context is complicated by issues of privacy, heterogeneity of the underlying data types, diverse semantic models, and the fundamentally complex nature of the health-care ecosystem. In this talk, I will discuss the ramifications of the underlying systems complexity of the health care system and how Grids and the associated concept of virtual organizations can provide solutions to the problems that result from this complexity. I will illustrate how Grid infrastructure can be applied within a number of clinical and research applications, including as part of the Biomedical Informatics Research Network (BIRN), a national scale medical information sharing infrastructure supported by the NIH.
April 2009 |
Joyce S. Chuang, Ph.D.
Graduate Student Researcher, University of California, San Diego
Jointly Presented by ICM & CIS
Hosted by Dr. Raimond Winslow & Dr. Michael Miller
"Noninvasive Phenotyping Of The Mouse Heart With Cardiac MRI"
Date: Thursday, April 9
Time: 1:00 pm EST
Location: Clark 314
Abstract:
The mouse model has become a powerful tool in cardiovascular disease research. Genetically engineered mice have been frequently used to study cardiomyopathies characterized by ventricular hypertrophy, dilatation, or diastolic/systolic dysfunction. Surgical methods such as aortic banding and coronary artery ligation can also induce pathological changes in the mouse heart. Despite the advantages of these models, studying the in vivo mouse heart is challenging due to the mouse's small heart size and fast heart rate.
Magnetic resonance imaging (MRI) is commonly used to obtain in vivo data of human or large animal hearts. Due to its high spatial and temporal resolution, MRI is ideal for studying geometry and function of the mouse heart. We have developed a MRI-based technique that accurately characterizes mouse cardiac geometry and function in 3D. Data from 2D cine anatomical MR images can be combined to generate a 3D ventricular model. The model allows for the measurement of different geometric parameters such as wall and cavity volumes. Deformation data from tagged MR images can also be incorporated to extract 3D strains throughout the heart. Our technique is useful for determining shape or strain patterns unique to different mouse models of cardiovascular disease.
January 2009 |
Terry Shen
Postdoctoral Candidate,
Karchin Lab
Johns Hopkins University
"Determining the feasibility and value of federated data integration combining logical and probabilistic inference for SNP annotation"
Date: Friday, January 16
Time: 3:30-5:00 pm EST
Location: CSEB 209
Abstract:
Most common and complex diseases are influenced at some level by variation in the genome. The future work of statistical geneticists, molecular biologists, and physician-scientists with interests in genetics or genomics must thus take genetics into consideration.
Research done in public health genetics, specifically in the area of single nucleotide polymorphisms (SNPs), is the first step to understanding human genetic variation. Functional uncertainty, volume of information, and cost-effectiveness result in the prioritization of SNPs to be an important research question. SNP Integration Tool (SNPit) is a data integration system tool that looks at all the possible predictors of functional SNPs and provides the user with integrated information and decision making capability. Determining the feasibility and value of SNPit with rules and probabilistic inference, thus, represents challenges from both the biological and biomedical informatics standpoint concerning how to represent, integrate, and conduct inference over disparate biological data sources.
I will discuss how to determine the feasibility and value of creating a federated integration system combining logical and probabilistic inference for functional SNP annotation. Preliminary studies include the prototype SNPit system which consolidates information on a variety of functional annotation predictors: everything ranging from gene expression, protein function, and evolutionary conservation through to known functional mutations.
August 2008 |
Hsiao-Fang Chou
Graduate Student,
Center for Imaging Science
Johns Hopkins University
"Smoothing Directional Vector Fields Using Dual Norms"
Date: Wednesday, August 27
Time: 12:00-1:00 EST
Location: Clark 110
Abstract:
We provide a new variational paradigm to measure the smoothness of unit vector fields on spatial domains, leading to new methods for smoothing and interpolating such datasets. Our point of view is to consider unit vector fields as linear forms acting on reproducing kernel Hilbert space of vector fields or tensors and work with the dual norm, leading to new variational problems and algorithms. We prove, in particular, that these variational problems are well-posed, in the sense that optimal solutions exist in the space of unit vector fields. Experimental results are based on synthetic data and DT-MRI (diffusion tensor magnetic resonance) datasets.
Hermenegild Arevalo
Graduate Student,
Institute for Computational Medicine
Johns Hopkins University
"Model of the infarcted canine heart predicts arrhythmia generation from specific cardiac geometry and scar distribution"
Date: Wednesday, August 13
Time: 12:00-1:00 EST
Location: Clark 110
Abstract:
Postinfarction reentrant ventricular tachyarrhythmias (VT) remain therapeutic challenges due to the difficulty in finding the location(s) of reentrant pathways arising from the infarct zone. Current clinical practice uses electrical activation mapping for pinpointing reentrant circuit location. The procedure is time consuming and is limited by the fact that reentrant circuits might be anchored to infarcted tissue located deep within the heart, away from recording electrodes. In this talk, we will present a novel technique that addresses this limitation by utilizing computer modeling to predict the location of reentrant circuits in an infarcted heart. High resolution magnetic resonance images and diffusion tensor images of an infarcted canine heart were used to construct a model that incorporates accurate geometry and fiber orientations. Image processing techniques were used to obtain accurate segmentation of the different regions of the infarction. Realistic ionic models were used to represent the electrophysiology within the different regions and simulations were performed to predict arrhythmia generation.
ICM Summer Fellows Symposium:
Date: Monday, August 4
Time: 2:00 EST
Location: Clark Hall, Room 110
Grace Tan
2008 Raj and Neera Singh Undergraduate Fellow
Institute for Computational Medicine
Johns Hopkins University
"Investigating Ventricular Tachycardia as a Result of Premature Stimuli in a Three-Dimensional Canine Cardiac Model of Heart Failure"
Abstract:
Experiments by Akar and Rosenbaum (2003) demonstrated increased arrhythmia vulnerability in canines induced to exhibit the characteristics of heart failure. In these hearts, the delivery of a single premature stimulus caused a conduction block at the subepicardial-midmyocardial cell interface, along the transmural wall of the heart. This was attributed to action potential duration (APD) prolongation in failing hearts, which is markedly heterogeneous across the transmural wall, and was characterized by disproportionate APD prolongation of midmyocardial cells. This subsequently led to the genesis of polymorphic ventricular tachycardia (PVT).
This project aims to replicate these experimental results in a three-dimensional wedge model of the failing canine heart. The wedge is built from diffusion tensor magnetic resonance imaging (DTMRI) data, experimental measurements of conduction velocity and connexin43 expression in failing and non-failing hearts, and utilizes the Greenstein-Winslow single-cell cardiac model.
Charlie Ouyang
2008 Raj and Neera Singh Undergraduate Fellow
Institute for Computational Medicine
Johns Hopkins University
"Fiber Tracing and Image Registration in MRI and DTMRI Data"
Abstract:
The Purkinje network is part of a specialized cardiac tissue network—the cardiac conduction system—and plays a role in the rapid propagation of impulses to the ventricular muscle. While the basic role of the Purkinje network is understood, some aspects of its behavior, for example its role in defibrillation shock response, remain unanswered.
The small size of the individual fibers and their positioning in the network makes direct observation of the network and its behavior difficult. Computer modeling of Purkinje network can provide understanding its structural implications with respect to electromechanical and biochemical properties of the heart. We develop a 3-D structural model of the Purkinje network by analyzing cardiac images. Magnetic Resonance Imaging (MRI) data were manually marked to form a ground truth fiber network to which the fiber network we generate by a nonlinear filter can be compared by taking symmetrized chamfer and Hausdorff distances between fiber tracts. We develop a block matching technique for image registration of Diffusion Tensor Magnetic Resonance Imaging (DTMRI) data. We were able to detect and recover transformations of the form x2 = Rx1 + t in both real and synthetic data.
July 2008 |
Francisco Sanchez Vega
Graduate Student,
Center for Imaging Science
Johns Hopkins University
"Probabilistic Graphical Models for Small-sample Network Inference"
Date: Wednesday, July 23
Time: 12:00-1:00 EST
Location: Clark 110
Abstract:
Probabilistic graphical models provide a powerful tool to efficiently model and analyze the conditional dependency relations among large sets of random variables. The correct identification of such relations plays a critical role when trying to obtain both quantitative and qualitative conclusions on the behavior of the modeled systems. This is more so in the context of small sample regimes that frequently characterize many problems in bioinformatics, such as the reconstruction of gene regulatory networks from microarray data. In this talk, we will present a general overview of some of the approaches that have dominated the literature during the last few years and we will sketch some new ideas regarding work currently in progress that is intended to overcome some of their limitations.
June 2008 |
Troy Anderson
Graduate Student,
Institute for Computational Medicine
Johns Hopkins University
"Application of Reverse Phase Protein Arrays to discovering the distribution of important proteins in the transmural proteome of the canine left ventricle"
Date: Wednesday, June 18
Time: 12:00-1:00 EST
Location: CSEB Auditorium
Abstract:
Reverse Phase Protein Arrays (RPPAs) are a relatively new method utilized in quantitative proteomics. Virtually all RPPA experiments are used to detect the concentration of a target protein present in relatively low abundance among a milieu of many higher abundance proteins extracted from a cell culture or tissue sample. RPPAs are unique in that they immobilize the entire protein extract on a nitrocellulose surface via robotic printing and probe the extract for the target protein with a specific antibody. These arrays will be utilized to measure the transmural distribution of proteins important in the excitation-contraction coupling process of the canine left ventricle. Due to the properties of RPPAs, it is hypothesized that we will be able to measure the distribution of proteins with greater statistical power and spatial resolution than previously achieved. Prior to the use of the arrays in the discovery phase, optimization and validation experiments needed to be carried out in order to ensure the worth of the collected data. In this talk, I will explore RPPAs in detail and share with you some of the results from optimization and validation of the arrays through spike-in protein studies and subsequent computational analysis. Furthermore, I will outline the future experiment which will collect transmural protein distribution data.
Dheeraj Singaraju
Graduate Student,
Center for Imaging Science
Johns Hopkins University
"Interactive Segmentation of Objects in Images via Minimization of Quadratic Energies on Graphs"
Date: Wednesday, June 11
Time: 12:00-1:00 EST
Location: CSEB Auditorium
Abstract:
Computer vision, in general, deals with the semantic interpretation of various objects present in an image. A task that is quintessential to computer as well as human vision, for obtaining such interpretations, is that of object segmentation. Essentially, it refers to the problem of finding within an image, the boundaries of certain objects of interest, or alternately, the regions that correspond to each of these objects.
Since images generally contain numerous objects that are further surrounded by clutter, it is often not possible to define a unique segmentation. In other words, the segmentation problem can be ill-posed when working in an unsupervised framework. Interactive algorithms allow the user to label a few pixels as either object or background, thereby making the segmentation problem well posed. Backed by this motivation, we propose to build an interactive system for the segmentation of objects in images. More specifically, our system will allow the user to mark representative scribbles that indicate certain objects of interest in the image, and consequently produce accurate segmentation of the objects of interest.
Our method proceeds by constructing a weighted combinatorial graph, such that each node in the graph corresponds to a pixel in the image. The edge weights of the graph are defined as measures of similarity between the image features of the nodes that they connect. The segmentation problem is then posed as the minimization of quadratic energies defined on this graph, subject to certain constraints defined by the user marked scribbles. We show that this framework finds equivalent constructions in electrical network theory. Such equivalent constructions can be used to generalize existing methods in order to introduce desired properties in the optimization scheme. In this talk, we will present preliminary work on using such equivalent electrical network constructions in order to devise general purpose object segmentation techniques.
April 2008 |
James Schwaber, Ph.D.
Director, Daniel Baugh Institute for Functional Genomics/Computational Biology
Department of Pathology, Anatomy and Cell Biology
Jefferson Medical College
Thomas Jefferson University
"Now, What is Life, Systems Biology of Homeodynamics: neuronal adaptive compensation"
Date: Wednesday, April 16
Time: 2:00-3:00 EST
Location: CSEB 320
Abstract:
What is Life?, Erwin Schrodinger's influential little book published in 1944, challenged biology and led to decades of gene-centric research that culminated in the mapping of the human genome. It asks “how can the events in... a living organism be accounted for by physics and chemistry?” In this 'gene's view' of life we have aimed to reduce diseases and functions to particular genes and molecules. It portrays life as serving the replication of genes and, at the limit in evolutionary psychology, brain functions such as behavior and cognition become merely modular products of genes selected in the Pleistocene. As a result anyone reading the popular biology literature will come away convinced that we have triumphed and that the essential problems have been solved by the 'genetic revolution'. The impression given is that we have reduced biological systems to simple, comprehensible genetic parts and that we are now on the threshold of being about to manipulate them as we do components in electronic devices.
This impression is false: the genes are not life. My seminar will analyze the emergence of post-genomic data that compels a very different view of the dimensions of the problem of the nature of life, and how that is driving development of the new field of systems biology. In these new data the genes are not controllers but participants in massive, multi-dimensional regulatory webs. This has the effect of standing "the selfish gene" on its head, and along with it much of the rationale for genetic determinism as the explanation for function. We now must understand life as a complex web or network, interacting across levels and in an ongoing process of profound adaptive interaction with its environment. Our task now is to point the way forward, just as Schrdinger successfully did for the issues of his day. As a framework in which to consider this challenge I will present my own work on the analysis of certain brain regulatory dysfunctions involved in hypertension.
Systems Biology of Homeodynamics: neuronal adaptive compensation
Recent reports indicate that the central nervous system plays a significant role in long term regulation of blood pressure, including the development and maintenance of hypertension, by baroreflex resetting to a higher mean arterial pressure set point. The mechanisms underlying these surprising but potentially very important cardiovascular homeodynamics are largely unknown. We hypothesize that the nucleus tractus solitarius (NTS), a major center mediating central and peripheral integration in cardiovascular control, adapts to peripheral blood pressure disturbance with a molecular remodeling that may contribute to alterations of regulatory function. In order to investigate this possibility we have mounted a systems-level study of the cardiovascular NTS response to hypertension. This study involves examination of system-wide transcriptional regulation, short- and longer-term signaling behavior, and the relation of these events to neuronal outputs including electrical behavior. We use a combination of high-throughput experimental methods and computational modeling, and both of these present unique challenges, for example in development of sufficient data quality, and in the use of very small samples. Our approach involves examining these processes together as a single cellular system. The behavior of this complex system involves dynamic interactions that are difficult to predict using qualitative reasoning and there is a need for experimentally validated computational modeling approaches at the systems level. These approaches are valuable in generation of hypotheses and provide a framework for the systematic comparison of data collected across experiments. These computational approaches also require particular focus on issues of broad significance in systems biology, such as the adequate representation of the complexity within a level of analysis, and the development of feedbacks or functional links between levels of analysis. The present seminar is combination of a progress report towards our scientific question of the basis for cardiovascular homeodynamics, our development of systems biology approaches making this feasible, and contemplation of the significant remaining challenges.
February 2008 |
Nathan Price, Ph.D.
Assistant Professor, Department of Chemical and Biomolecular Engineering Institute for Genomic Biology
University of Illinois, Urbana-Champaign
"Systems Biology in Medicine: Applications to Anemia, Diabetes, and Cancer"
Date: Friday, February 8
Time: 1:00-2:00 EST
Location: Clark Hall 314
Abstract:
Systems biology - the intersection of high-performance computing with high-throughput experimental technologies to drive biological discovery - has the potential to transform the practice of medicine to a more predictive and personalized discipline. These technologies enable the reconstruction and modeling of large-scale biochemical networks in human systems that can aid in the understanding, and eventual control, of disease phenotypes. In addition, disease-perturbed biomolecular networks result in altered levels of gene and protein expression that necessarily leave molecular fingerprints that can be identified for accurate disease diagnosis to aid in the selection of appropriate therapies for the patient. Systems approaches to human disease will be illustrated in three cases: hemolytic anemia, diabetes, and cancer.
November 2007 |
Rui Chang
Institute of Informatics, Department of Computer Science, Technical University of Munich, Germany
"Advanced Probabilistic Network Modeling with Qualitative Prior Knowledge and Application in Quantitative Inferring Multi-scale Molecular Interactions Network"
Date: Tuesday, November 13
Time: 2:15-3:15 EST
Location: Clark Hall 314
Abstract:
In this talk, I will propose unprecedented solutions to the challenges in Bayesian network learning, namely, how to construct prior distribution over structure and parameter space from prevalent amount of preexisting qualitative information in science and industrial domain within an unified framework as well as to the tough question how qualitative statements about relationship between domain entities can be transformed to yield quantitative predictive models, able to perform probabilistic inference and reasoning. To this end, we will only consider the statistics and uncertainty presented by prior information, i.e. we utilize solely qualitative prior information in our study and therefore, no quantitative data information is available to shield our insights in the function and effects of prior knowledge in probabilistic modeling with Bayesian networks.
September 2007 |
Dr. Youn-Seon Lee
Postdoctoral Candidate
"Exploring Ca2+ release mechanisms in cardiac cells"
Date: Wednesday, September 12
Time: 2:00 PM EDT
Location: Clark Hall 110
Abstract:
Ca2+-induced Ca2+ release (CICR) is widely accepted as the principal mechanism linking between electrical excitation and mechanical contraction in cardiac cells. However, Ca2+ release mechanisms that determine the fraction of sarcoplasmic reticulum (SR) Ca2+ release are not well understood. Using mathematical models we explored two possible mechanisms that may lead to nonlinear fractional SR Ca2+ release. First, a mathematical model, based on the one-dimensional reaction-diffusion model of Keizer-Smith 1998, showed that alternating patterns of Ca2+ wave propagation can occur with periodic Ca2+ stimulation under conditions of Ca2+ overload. From this model we speculate that such behavior results from a delay in recovery from refractoriness of RyR channel kinetics. Second, we developed a kinetic model of RyR that has three binding sites, two cytosolic sites for Ca2+ activation and inactivation, and one SR luminal site for calsequestrin (CSQ) binding. The RyR kinetic model was incorporated into a local CICR model that has both a diadic space and junctional SR (jSR). This model suggests that CSQ has an inhibitory effect on RyR gating at low jSR load as more CSQs bind to RyR. This local CICR model also produces a nonlinear fractional relation of jSR Ca2+ release on jSR load. These putative Ca2+ release mechanisms currently are being investigated experimentally in canine cardiac Purkinje cells.
February 2007 |
Nilay Roy, Ph.D.
Postdoctoral Candidate
Department of Chemistry, Brandeis University
"Using large scale computer simulations for solving intractable problems"
Date: Tuesday, February 20
Time: 11:00 AM EST
Location: Clark Hall 110
Abstract:
In this talk I will briefly present several examples where intractable problems are solved using a combination of simulational methodologies. These problems include that of polymer property prediction and optimization and polymer blend design, identifying the pathway of polymorphic transformations in animo-acids, drugs, proteins and other small molecules, and simulating ion channels in large lipid bilayers. The combination of techniques used includes Neural Networks, Genetic Algorithms, Markov Chains, Molecular Dynamics and semi-explicit methods. An outline will then be presented on how these techniques can be extended and modified to identify potential cancer genes from the genome, and to model the effects of these mutations on protein structure, activity and function.
Parminder Mankoo, Ph.D.
Postdoctoral Candidate
Department of Chemistry, Boston University
"How computer simulations help us understand recognition of ligands in proteins, vibrational Stark effect and infrared spectroscopy"
Date: Tuesday, February 13
Time: 2:00-3:00 EST
Location: Clark Hall 110
Abstract:
Protein structure-function relationship is one of the cornerstones of modern biology, and holds the key to decipher many problems in physiology and medicine. For example, the primary function of myoglobin (Mb) is to transport oxygen in organisms. Carbon monoxide is a poison of Mb, and the oxygenated form of Mb prevents nitric oxide from inhibiting cytochrome c oxidase, making Mb a protector of cellular respiration. Experiments have demonstrated that the structures of myoglobin enable the protein to function and to distinguish among oxygen, carbon monoxide and nitric oxide. To understand protein structure-function relationships and spectroscopy at the molecular level, computer simulations are often helpful and sometimes necessary. However, current simulation methods are often time- and labor-intensive, or they perform poorly in computing spectroscopic properties and protein-ligand binding energies.
To counter these difficulties, we have constructed a new model for computer simulations based on classical molecular electrostatics that includes, crucially, a proper treatment of polarization of atoms. When applied to myoglobin interacting with a ligand, our model successfully predicts the ligand-heme binding and bending energies, captures the vibrational Stark effect of ligands and discriminates the myoglobin binding affinities among carbon monoxide, nitric oxide and oxygen. Our model explains these properties and the preferential binding of one atom of the ligand to heme in intuitive electrostatic terms. Our approach demonstrates the power of computer simulations in understanding biology, and can be readily extended to investigate other problems of medical importance such as drug-receptor binding, amyloid formation and the effects of point mutations in cancer-related proteins, where electrostatics may play a significant role.
January 2007 |
Nicole Leahy, Ph.D.
Postdoctoral Candidate
Department of Epidemiology and Preventive Medicine, University of Maryland
"Applications of genetic algorithms and graph theory to genomic problems"
Date: Tuesday, January 30
Time: 2:00-3:00 EST
Location: Clark Hall 110
Abstract:
In the genomic era, a challenge is predicting the effect of single nucleotide polymorphisms (SNPs) without extensive wetlab testing of each new variant. The computational methods are applied to this problem to avoid expensive and time consuming wetlab experiments. Evolutionary algorithms and graph theory are sources of such methods. An evolutionary algorithm (EA) attempts to find a workable solution by successive rounds of testing and "breeding" of a population of solutions. The choice of representation of a solution depends on the nature of the problem under study. Logical trees and lookup tables provide possible representations which lend themselves to EA. Graph theory can be used to map out an evolutionary "space" which may lead to insights about patterns of deleterious mutations.
October 2006 |
Shankar Subramaniam, Ph.D.
Professor, Departments of Bioengineering, Chemistry and Biochemistry
University of California at San Diego
"Systems Biology of Macrophages"
Date: Wednesday, October 25
Time: 1:00-2:00 EST
Location: Clark Hall 110
Simulcast to Talbot Library
Abstract:
Cells and tissues function in context. Under a given growth or survival medium they perform tasks, replicate and die. Given a stimulus they respond by invoking myriad biomolecular networks that result in a specified cellular outcome. At any given instant it can be argued that the cell is in a “state” defined by its components - their concentrations and locations, the interactions between components - that are modulated in space and time, and the complex circuitry - that involves a large number of interacting networks and a snapshot of the dynamical processes - such as gene expression, cell cycle, transport of components, etc. At present, we can measure, using high and low throughput methods, several cellular components in a context-dependent manner and obtain a partial picture of cellular networks and dynamical processes. Are these measurements sufficient to answer important biological questions and help reconstruct a systems-level of understanding of a mammalian cell? This talk will address strategies developed to address this question and demonstrate the power of integration of diverse cellular data for answering interesting biological questions in macrophages. We will use this systems biology approach to address the following questions:
• How good are macrophage cell lines in addressing phenotypic biology of primary macrophages?
• How can we combine proteomic and other cellular measurements to characterize the repertoire of upstream signaling networks invoked by macrophages?
• How do signals associated with inflammatory molecules regulate gene transcription in macrophages?
• How does lipid signaling influence the proteomic pathways associated with Toll receptor pathways?
• How can we combine heterogeneous data to quantitatively decipher cross-talk in macrophage signaling?
• How do designed knockdowns of proteins influence cellular phenotypes?
April 2006 |
Gregory Smith, Ph.D.
Professor, Department of Applied Science
The College of William and Mary
"Stochastic Modeling of Local and Global Intracellular Calcium Dynamics"
Date: Wednesday, April 12
Time: 11:00-12:00 EST
Location: Clark Hall Room 110
Abstract:
Although there is consensus that calcium (Ca) puffs and sparks arise from the cooperative action of multiple intracellular Ca channels, the precise relationship between single-channel kinetics and the collective phenomena of stochastic Ca excitability and oscillations is not well understood. Here we present and analyze a stochastic automata network model of instantaneously coupled Ca-regulated Ca channels that gives insight into how the stochastic dynamics of an individual Ca release site depends on channel density and the presence or absence of Ca inactivation. The relationship between such stochastic Ca release site dynamics and global Ca responses will then be discussed in the context of a novel probability density approach to modeling whole cell Ca dynamics. The method involves coupling ODEs for the bulk cytosolic and ER [Ca] to advection-reaction equations for the probability density of the [Ca] in cytosolic and lumenal domains associated with each channel and conditioned on channel state. The probability density approach is computationally more efficient than explicitly spatial Monte Carlo simulations and the representation of local Ca signals inherent in the probability density approach is more realistic than conventional stochastic compartmental models.
Rafael Ramirez, Ph.D.
Heart & Stroke/Richard Lewar Centre of Excellence
University of Toronto
"Intracellular [Na+] modulates synergism between Na+/Ca2+ exchanger and Ca2+ current in cardiac excitation-contraction coupling during an action potential"
Date: Wednesday, April 19
Time: 11:00-12:00 EST
Location: Clark Hall Room 110
Abstract:
Contraction of cardiac myocytes is initiated by Ca2+ influx through L-type Ca channels, which induces Ca2+ release from the sarcoplasmic reticulum (SR). The contribution of Ca2+ entry via the Na+/Ca2+ exchanger (NCX) to triggering Ca2+ release during excitation-contraction coupling (ECC) during an action potential (AP) remains uncertain. In order to isolate the contribution of NCX to SR Ca2+ release, independent of its effects on SR Ca2+ load, Ca2+ release was determined by recording Ca2+ spikes using confocal microscopy of patch-clamped rat ventricular myocytes with [Ca2+]i fixed at 150 nmol/L. In response to AP clamps, normalized Ca2+ spike amplitudes (ΔF/F0) increased (P<0.005) sigmoidally from 0.40±0.08 to 0.80±0.06 as [Na+]i was elevated from 0 to 20 mmol/L with an EC50 of 9.08±0.97 mmol/L. The [Na+]i dependent enhancement of SR Ca2+ release was independent of INa or SR Ca2+ load. However, NCX inhibition using either 5 μmol/L KB-R7943 or 30 μmol/L XIP reduced (P<0.05) ΔF/F0 amplitudes in myocytes with 20 mmol/L [Na+]i but not with 5 mmol/L [Na+]i. Since ICa,L inhibition with 50 mmol/L Cd2+ totally abolished Ca2+ spikes, our results demonstrate that, during a cardiac AP at elevated [Na+]i, NCX enhances SR Ca2+ release, by synergistically increasing the efficiency of ICa,L-mediated ECC. Additionally, the slope of the initial repolarization phase of the cardiac AP is a key regulator of ECC. We hypothesize that this synergy between NCX and ICa,L is responsible for AP dependent modulation of cardiac ECC. In voltage-clamped rat ventricular myocytes, we measured Ca2+ spikes triggered by a family of APs with varying slopes of initial repolarization. In myocytes containing 20 mmol/L [Na+]i, ΔF/F0 displayed a biphasic relationship with AP duration at 50% repolarization (APD50), rising initially then decreasing as APD50 of the triggering AP increased from 4 to 52 ms. Maximal Ca2+ release flux was achieved when APD50 was 16 ms (ΔF/F0=0.77±0.06, n=17). In the presence of XIP the APD-dependence of Ca2+ spikes was blunted, ΔF/F0 decreased (P<0.02) to 0.60±0.06 at 16 ms APD50. In myocytes containing 5 mmol/L [Na+]i, the APD-dependence was also blunted (ΔF/F0=0.53±0.06, at 9 ms APD50, n=16) and was unaffected by XIP. We conclude that the biphasic enhancement of SR Ca2+ release with prolongation of APD is achieved through functional synergy between NCX and L-type Ca2+ channel trigger sources of CICR.
Pranay Goel, Ph.D.
Postdoctoral Fellow, Mathematical Biosciences Institute
Ohio State University
"A Three-Dimensional Model of Calcium Clearance in a Cardiac Myocyte During Excitation-Contraction Coupling"
Date: Thursday, April 20
Time: 11:00-12:00 EST
Location: Clark Hall Room 110
Abstract:
I shall describe a minimal, spatial model of calcium transients in a cardiac myocyte underlying E-C coupling. Calcium and buffer dynamics in the cytosol and the sarcoplasmic reticulum are described by reaction-diffusion systems solved on a three-dimensional compartmental geometry that is taken to approximate the sarcomere. The model incorporates fluxes due to L-type calcium channels, serca pumps, ryanodine receptors and sodium-calcium exchange. In particular, we introduce a bidomain formulation of the network SR and cytosol that can be regarded as a coarse-graining over the reticular geometry of the network; I shall point out how this bidomain model is formally derived using homogenization theory. The model reproduces various marcoscopic fluxes that are consistent with experimental data.
Mark Ellisman, Ph.D.
Director, Center for Research in Biological Systems (CRBS); Director, The National Center for Microscopy and Imaging Research
University of California San Diego
"Multi-scale Imaging of the Nervous System: Where's the Dark Matter?"
Date: Thursday, April 27
Time: 3:00-4:00 EST
Location: Clark Hall Room 110
Abstract:
We are evolving a shared infrastructure that allows for mapping molecular and cellular brain anatomy in the context of a shared multi-scale mouse brain atlas system, the Cell-Centered Database (CCDB). Complementary to these neuroinformatics activities, we have developed new molecular labeling methods compatible with advanced ultra-wide field laser-scanning light microscopy and multi-resolution 3-dimensional electron microscopy. The informatics framework is facilitating cooperative work by teams of scientists engaged in collaborations aimed to deliver new fundamental understanding of structures on the scale of 1 nm3 to 10's of µm3, a dimensional range that encompasses macromolecular complexes, organelles, and multi-component structures like synapses and the cellular interactions responsible for the functional organization of the nervous system.
March 2006 |
Harlan Robins, Ph.D.
Center for Systems Biology
Institute for Advanced Study
"Extending the Principle of Maximum Entropy to Find Motifs in the Genome"
Date: Tuesday, March 14
Time: 11:00-12:15 EST
Location: Clark Hall Room 110
Abstract:
The degeneracy of codons allows a multitude of possible sequences to code for the same protein. Hidden within the particular choice of sequence for each organism are over a hundred previously undiscovered biologically significant short (length 2-7) oligonucleotides. We present an information-theoretic algorithm that finds these novel signals. Applying this algorithm to the 209 sequenced bacterial genomes in the NCBI database, we determine a set of oligonucleotides for each bacterium which uniquely characterizes the organism. Additionally, applying the algorithm to the human and HIV genome, we find evidence for particular binding site that is involved in the HIV life cycle. The methods developed here can be readily extended to other problems in bioinformatics.
Tom Shannon, Ph.D.
Department of Molecular Biophysics & Physiology
Rush University
"Mathematical and Experimental Studies Addressing the Role of Diastolic Ca Release in Cardiac Excitation-Contraction Coupling"
Date: Wednesday, March 15
Time: 3:00-4:00 EST
Location: Clark Hall Room 110
Abstract:
Cardiac excitation-contraction coupling is initialized by the release of Ca from the sarcoplasmic reticulum (SR) in response to a sudden increase in local cytosolic [Ca] ([Ca]i) within the junctional cleft. We have tested the hypothesis that functional ryanodine receptor (RyR) regulation plays a major role in the regulation of myocyte Ca. A mathematical model with unique characteristics was used to simulate Ca homeostasis. Specifically, the model was designed to accurately represent the SR [Ca]-dependence of release from a variety of experimentally produced data sets which I will present. The simulated data for altered RyR Ca sensitivity demonstrated a regulatory feedback loop that resulted in the same release at lower [Ca]SR. This suggests that the primary role of myocyte RyR regulation may be to decrease SR [Ca] without decreasing the size of the [Ca]i transient. The model results suggest that this action moderates the increased SR [Ca] observed with adrenergic stimulation and may keep the [Ca]SR below the threshold for delayed afterdepolarizations and arrhythmia. However, increased Ca affinity of the RyR increased the probability of delayed afterdepolarizations when heart failure was simulated. We conclude that RyR regulation may play a role in preventing arrhythmias in healthy myocytes but that the same regulation may have the opposite effect in chronic heart failure.
Itsik Pe'er, Ph.D.
Broad Institute, Whitehead Institute and Massachusetts General Hospital
"Data, technology and populations for genomewide association studies"
Date: Thursday, March 16
Time: 11:00-12:15 EST
Location: Clark Hall Room 110
Abstract:
The pervasive effect of genetic variation on medically important phenotypes provides a means for dissecting their underlying mechanisms by identifying variants that are associated with traits of interest. Current trends in human genetics now facilitate, for the first time, pursuing this potential by execution of large scale studies that scan the entire genome for potentially associated variants. Specifically, the talk will present
-The International HapMap Project, a data resource we participated in developing to enable genomewide association studies, and what our analyses of these data tell us about human variation.
-The current generation of SNP array technology, and how computation and statistics improvements allow it to cover the majority of common human variants.
-The tale of an isolated population in Micronesia, where we show association scans are more promising than elsewhere, though we expose practical complexities of real data and the computational challenges they present.
Some of the research presented was performed as part of the International HapMap Analysis Team, or in collaborations with Affymetrix Inc. and the Friedman lab at Rockefeller University.
Rachel Karchin, Ph.D.
Andrej Sali Lab
University of California, San Francisco
"Making sense of germline mutants that impact human health"
Date: Tuesday, March 21
Time: 11:00-12:15 EST
Location: Clark Hall Room 110
Abstract:
Germ-line DNA variation that results in a single amino-acid residue change in the protein product of a gene (missense mutant) may have a major impact on an individual's susceptibility to disease and sensitivity to drugs. Many such variants occur at very low population frequencies, thus case/control
and familial cosegregation studies are not sufficiently powered to discriminate between those which are pathogenic/high clinical significance vs. neutral/low clinical significance. A promising alternative approach is to integrate information derived from computational biology with clinical patient data and functional studies.
I will describe work that applies protein homology modeling, sequence analysis, and machine learning to predict and rationalize the impact of missense mutations on protein stability and function. These predictions can complement information from patient pedigrees, loss of heterozygosity studies of tumor tissue, and help make sense of the results of functional assays. I will also discuss how the process can be automated and applied to large-scale datasets.
Felim MacGabhann
Department of Biomedical Engineering
Johns Hopkins University
"What do cancer and heart disease have in common?
Designing therapeutic strategies by modeling angiogenesis"
Date: Thursday, March 23
Time: 11:00-12:15 EST
Location: Clark Hall Room 110
Abstract:
Tissue does not survive without adequate vasculature; this is as true of growing tumors as it is of ischemic heart muscle. Angiogenesis (or neovascularization), the outgrowth of new vessels from the preexisting vasculature, is essential to organ development, and is involved in several physiological processes in the adult, including wound healing. Many diseases may be treated by either promoting or inhibiting the angiogenic process. In the first category, both coronary and peripheral limb ischemia may be alleviated by increasing the overall vascularization of the tissue. In the second category, tumor growth may be slowed by inhibiting angiogenesis; later stage cancers may be more effectively chemotargeted when the tumor vessels are not growing erratically; and metastatic events may be decreased by reducing the tumor’s vascular supply. Other diseases which are vasculature-dependent and can therefore be targeted with angiogenesis inhibitors include diabetic retinopathy, age-related macular degeneration and arthritis.
The principal cytokine implicated in the induction of angiogenesis is Vascular Endothelial Growth Factor (VEGF). Released by hypoxic tissue, it binds receptors on the blood vessels and guides the growth of sprouts towards the hypovascularized area. We have constructed the first molecularly-detailed models of the interactions of VEGF with its receptors and its transport through various tissues. The models, validated using all available experimental evidence, are used to predict and compare the efficacy of cell-, protein-, and gene-based therapies. In this talk, I will focus on two applications of these models. First, strategies for inhibiting VEGF signaling in tumors by targeting the VEGF co-receptor Neuropilin-1. Second, pro-angiogenic therapies for alleviating hypoxia in peripheral artery disease.
February 2006 |
Jeremy Rice , Ph.D.
Functional Genomics and Systems Biology
IBM T.J. Watson Research Center
"A spatially-detailed model of actin and myosin interaction in the cardiac myofilament "
Date: Thursday, February 2
Time: 3:00 P.M. EST
Location: Clark 110
Abstract:
The availability of increased computing power will make
new classes of biological models possible. One such class
will be spatially explicit representations of subcellular
machinery. We have developed such a model of the interaction
of actin and myosin within one pair of thick and thin
filaments in the cardiac sarcomere. We have chosen to
model the smallest repeating unit that maintains fundamental
spatial interactions thought to produce important and
characteristic cardiac muscle responses at cell, tissue
and whole heart levels. The model discussed differs from
existing models by having spatially explicit representations
of actin, myosin and regulatory protein interactions including
extensible links to capture filament compliances. Sample
results are compared with experimental characterizations
of real muscle. The model recapitulates a wide range of
complex, non-linear behaviors and hence provides a plausible
and quantitative explanation for several unexplained phenomena
in cardiac muscle. The execution of the model demands
Monte Carlo-based simulations of Markov representations
of Ca regulation and actin-myosin interactions with strain-dependent
rates. In addition, solution of the compliant realignment
of the thick and thin filament binding sites requires
an iterative solution technique that demands substantial
computational resources. The model is suitable to serve
as a basis for larger scale simulations.
Yi Jiang, Ph.D.
Theoretical
Division
Los Alamos National Laboratory
"A Multiscale
Model for Tumor Growth"
Date: Tuesday, February 7, 2006
Time: 1:30 P.M.-3:00 P.M. EST
Location: Clark Hall 110, Simulcast to Talbot Library, 709 Traylor Bldg
Abstract:
Cancer has become the leading cause of death for Americans. The development of prognostic tools could have immediate impact on the lives of millions of cancer patients. We have developed a multiscale model that integrates a cellular model for cell dynamics (cell growth, division, death, and cell adhesion), an intracellular protein regulatory network for cell cycle control, and extracellular reaction-diffusion chemical dynamics. This model has produced tumor growth dynamics that agree with tumor spheroid experiments, and generated a few hypotheses on tumor biology that can be tested by experiment. The model has the potential to become a comprehensive and predictive model for tumor development and therapy based on quantitative experiments.
Hin Hark Gan, Ph.D.
Research Assistant Professor, Department of Chemistry
New York University
"Computational Approaches to RNA Genomics and Design"
Date: Tuesday, February 14, 2006
Time: 11:00 A.M.-12:00 P.M. EST
Location: Clark Hall 110
Abstract:
RNAs are wonderfully versatile molecules. Recent discoveries reveal that RNA molecules play vital roles in cellular function and disease, and have applications in functional genomics and the design of molecular sensors. We will describe graph theory-based and other computational methods to uncover natural non-coding RNAs in genomes, isolate synthetic functional RNAs from large sequence pools, and predict novel antibiotic targets on ribosomal RNAs. These efforts are conducted in conjunction with multidisciplinary experimental collaborators. Our graph theory description of RNA molecules predicts the diversity, abundance, and organization of the RNA structure universe. The predicted RNA-like structures provide a unifying framework for identifying functional RNAs. Specifically, we exploit such RNA-like structures to identify several novel non-coding RNAs in the human chromosome region associated with Prader-Willi and Angelman syndromes. As a complementary strategy, we use synthetic functional RNA motifs to identify their genomic counterparts which may have cellular roles similar to riboswitches. In parallel to genome searches, we have developed computational approaches for enhancing in vitro selection technology – an experimental technique for isolating functional RNAs from large sequence pools – by engineering structurally diverse RNA pools. This work combines pool synthesis concepts, graph theory, and analysis of RNA sequence/structure space to design sequence pools. In addition, we employ thermodynamic methods for siRNA design to predict known and novel antibiotic targets on ribosomal RNAs. Thus, our computational approaches offer tools for advancing RNA genomics, identifying RNAs implicated in diseases, improving in vitro selection technology, and predicting antibiotic-binding targets.
Gabriel Kreiman, Ph.D.
Whiteman Science Fellow, Dept. of Brain and Cognitive
Science and Computation and Systems Biology Initiative
Massachusetts Institute of Technology
"Combining machine learning and computational models to study biological codes"
Date: Tuesday, February 21, 2006
Time: 11:00 A.M.-12:00 P.M. EST
Location: Clark Hall 110
No podcast available
Abstract:
Codes are prevalent throughout biological systems
(as in the map from the information in DNA to the sequence
of aminoacids in proteins). A particularly appealing aspect
of biological codes is that they exhibit some of the main
defining properties of Systems including complex emergent
behaviors, non-linear interactions, adaptability and robustness
to perturbations. Pathological situations including cancer,
epilepsy and brain disorders arise from the properties
of the system as a whole and therefore quantitative modeling
of these systems will play a crucial role in unraveling
the mysteries of how these systems work or fail to work.
In this talk I will show how combining machine learning
techniques and computational models can help us investigate
two types of biological codes: (i) the mechanisms that
control gene expression and processing and (ii) how circuits
of neurons in the brain represent information. I will
show how computational algorithms can help us understand
the organization of cis regulatory elements in eukaryotic
genomes and I will illustrate the progress made by using
computational models towards characterizing and understanding
the function of neuronal circuits.
January
2006 |
Natalia
Trayanova, Ph.D.
Professor, Department of Biomedical Engineering
Director, Computational Cardiac Electrophysiology Laboratory
Tulane University
"Virtual Reality in the Race against Sudden Cardiac Death"
Date: Tuesday, January 17, 2006
Time: 11:00 A.M.-12:00 P.M. EST
Location: Clark Hall 110, Simulcast to Talbot Library, 709 Traylor Bldg
Abstract:
This talk presents an overview of the research conducted at the Computational Cardiac Electrophysiology Laboratory at the Department of Biomedical Engineering, Tulane University, of which Dr. Trayanova is the Director. The projects range from cellular-level to multi-scale whole-organ simulations aimed at providing mechanistic insight into various problems in cardiac arrhythmogenesis and anti-arrhythmia therapies. Projects include arrhythmogenesis under the conditions of ischemia phase 1A, mechano-electric feedback in the heart, studies on the mechanisms for ventricular defibrillation, the role of fibroblasts in conduction and defibrillation, and arrhythmia induction due to electroporation caused by strong shocks in the heart.
Alan Garfinkel, Ph.D.
Professor, Medicine and Physiological Science , UCLA
"Mechanisms of Atrial and Ventricular Fibrillation"
Date: Wednesday, January 25, 2006
Time: 3:00-4:00 P.M. EST
Location: Clark Hall Room 110, Simulcast
to Talbot Library, 709 Traylor Bldg
Abstract:
In the healthy
heart, the wave of contraction that pumps the blood is
created by a wave of electrical depolarization that passes
from cell to cell through the myocardium. But in fibrillation,
the depolarization wave breaks up into “electrical
turbulence,” and coherent contraction is lost.
What causes this breakup? Cardiologists had long assumed
that the cause of wavebreak must be external ‘wavebreakers,’
that is, areas of reduced conduction due to infarction,
ischemia, fibrosis, etc. In contrast, physicists have
shown that it is possible for a wave to break up due solely
to the internal dynamics of wave propagation. Since the
dynamics of wavefront formation and propagation depend
in turn on ion channel dynamics within the cardiac myocytes,
it follows that modifications of these ion channel dynamics
could potentially stabilize the wave of propagation and
prevent the breakup into fibrillation.
Our group at UCLA and Cedars-Sinai has pursued this strategy,
using mathematical and physical theory together with very
large scale models of cardiac propagation in realistic
models of the atria and ventricles (a “Virtual Heart”).
We have found that it is indeed possible to prevent the
breakup into fibrillation by using drugs to modify ion-channel
kinetics in the myocyte.
This talk will emphasize the mathematical and physical
theory of wave conduction and wavebreak. We will present
several equations, for wave propagation and for intracellular
calcium dynamics, that play critical roles in the breakup
that is fibrillation.
Our initial work on wave stability theory has since been
supported by experiments in animal hearts, which confirm
that modifying cell dynamics can serve as a potent anti-fibrillatory
strategy, giving rise to new types of pharmacological
and genetic interventions against atrial and ventricular
fibrillation.
Joel Saltz, Ph.D.
Professor and Chair of the Department of Biomedical
Informatics, Ohio State University
"The Data Driven Challenge of Translational Research"
Date: Thursday, January 26, 2006
Time: 3:00-4:00 P.M. EST
Location: Clark Hall Room 110
Abstract:
Translational biomedical research
is a data driven challenge. Ability to carry out complex
translational studies will rely on effectively tackling
challenges posed by information heterogeneity, data coordination
and data size. Generating integrated views of biomedical
phenomena can involve the need to synthesize information
from many high throughput
information sources. These sources can include multiple
types of high throughput molecular data and multiple imaging
modalities. Translational studies often involve coordinated
efforts at multiple sites leading to the need to share
detailed experimental data and the need to coordinate
work. Finally, detailed understanding of biomedical phenomena
will
increasingly involve the need to analyze very large high
resolution spatio-temporal datasets.
We will present an overview of middleware and algorithms
designed to respond to challenges posed by heterogeneity,
coordination and data size. We will outline the architectural
roles played systematic metadata management mechanisms,
and by mechanisms that provide an abstract view of combined
information stored in filesystems, XML databases and in
relational databases. Finally, we will motivate this talk
by a variety of examples that arise in exploring the morphological
impact of Rb gene knockouts, in grid based CAD image analysis
and in caBIG based translational research scenarios.
December
2005 |
Herbert Levine, Ph.D.
Professor, Department of Physics
"Excitation-contraction coupling gain and cooperativity
of the Ryanodine receptor: a modeling approach"
Date: Wednesday, December 14, 2005
Time: 3:00 - 4:00 P.M. EST
Location: Clark Hall Room 110
No podcast available
Abstract:
Intracellular calcium dynamics
is a prime example of a stochastic excitable system. One
therefore expects generically to see the spontaneous generation
of localized calcium events (“sparks”) and/or
extended calcium waves, depending on the amplification
kinetics and the spatial coupling between release units.
Specific calcium-signaling systems, for example cardiac
excitation-contraction (EC) coupling, need to be engineered
so maximize functionality given this dynamical context.
This talk will review some of the basic phenomenology
of stochastic excitable dynamics and then turn to one
proposed mechanism by which the EC system maintains gain
even while preventing excess leak through spontaneous
sparks. Specifically, we show by means of a simplified
model that a proper balance between independent and coupled
gating of the four Ryanodine subunits can accomplish this
task. Experimentally, this balance appears to be regulated
by the binding of FKBP; hence our model sheds light on
possible diseased states with altered FKBP binding activity.
November
2005 |
Chris Johnson, Ph.D.
Director, Scientific Computing
and Imaging Institute, University of Utah
"Computing the Future of Biomedicine"
Date: Tuesday, November 8, 2005
Time: 1:00 - 2:00 P.M. EST
Location: Clark Hall Room 110
No podcast available
Abstract:
Computers have changed the way we live, work, and even recreate. Now, they are transforming how we think about and treat human disease. Advanced techniques in biomedical computing, imaging, and visualization are already changing the face of biology and medicine in both research and clinical practice. These techniques have the potential to provide comprehensive models and views of the human body in unprecedented depth and detail. As a result, biomedical computing and visualization will help produce exciting new biomedical scientific discoveries and clinical treatments. In this talk, I will discuss the state of the art in biomedical computing, medical imaging, and visualization research and present examples of their vital roles in cardiology, neuroscience, neurosurgery, and radiology.
Saleet Jafri , Ph.D.
Associate Professor, School of
Computational Sciences, George Mason University
"A Computational Study of Calcium Sparks in Cardiac
Cell"
Date: Wednesday, November 9, 2005
Time: 3:00 - 4:00 P.M. EST
Location: Clark Hall Room 110
No podcast available
Abstract:
The primary function of the heart is to pump blood to and from the body by synchronized contraction and relaxation of the heart muscle. The process which enables the heart to do so is called excitation-contraction coupling. Excitation refers to the depolarization of the cell membrane that encapsulates heart muscles. During excitation, the opening of voltage-gated L-type calcium channels triggers calcium release from clusters of ryanodine receptors (calcium induced calcium release channels) in the sarcoplasmic reticulum (the intracellular calcium stores). Release from an individual cluster is called local and elementary release events are called calcium sparks. The calcium released during sparks sums up to form the global calcium transient that binds to the contractile proteins and initiates the contraction. Any imbalances in calcium levels can lead to contractile dysfunction, cardiac arrhythmia and heart failure. Therefore, due to the crucial role of calcium EC coupling, it is important to understand calcium dynamics at the elementary level of calcium sparks.
We have developed a three-dimensional model of calcium sparks to analyze the basic mechanisms of cardiac muscle contraction through integration of detailed biophysics and microanatomy including the effects of experimental measurement techniques on the observed results. This talk will be focused on the spread of calcium sparks and the existing discrepancy in spark width (full width, half-maximum, FWHM) between experimental results (~2.0 um) and computational models (1.0-1.2 um) under physiological conditions. It will be discussed how simplified assumptions about the morphology of the organelles and limitations of measuring techniques can significantly affect calcium spark spread of computational models.
October
2005 |
David Mumford, Ph.D.
University Professor, Division of Applied Mathematics, Brown University
"What's an Infinite Dimensional Manifold and How Can it be Useful in Anatomy?"
Joint Seminar with the Center for Imaging Science
Date: Thursday, October 20, 2005
Time: 2:00 - 3:00 PM EST
Location: 110 Clark Hall
July
2005 |
Siamak Ardekani
Graduate Student, Biomedical Engineering, University of
California at Los Angeles
"Diffusion MRI Brain Atlas at 3.0T"
Date: Wednesday, July 27, 2005
Time: 3:00 - 4:00 PM EST
Location: 110 Clark Hall
No podcast available
Abstract:
Quantitative measurements of brain diffusion parameters
(apparent diffusion coefficient [ADC] and fractional anisotropy
[FA]) provide new insight compared to conventional MRI.
Underlying structural changes at the cellular level can
give rise to changes in diffusion parameters that is observable
at image spatial resolutions. A standard set of diffusion
parametric values (and their spatial distribution) provides
a basis for identifying and monitoring abnormality in
a given patient relative to the reference. However, the
inherent normal variations present in normal human brains
will complicate distinguishing abnormalities from normal
variants. Clearly, a single brain cannot accurately represent
the parameter variance in the normal population. Probabilistic
atlas methods, addresses this challenge by providing methods
to capture the population variability. On the other hand,
intensity-based models are focusing on reconstructing
an average representation of structural anatomy by averaging
over multiple MRI scans.
In this work we utilize affine and free form transformations to successfully create a minimally distorted average morphometric atlas of diffusion parameters (FA, ADC) acquired at 3.0 T. This atlas converges towards the centroid of population data set and reflects the variability in diffusion parameters for the ten subjects on a voxel-by-voxel basis. The atlas can be utilized as a tool that characterizes morphological or functional information in order to distinguish abnormal anatomical and functional variations. A pre-requisite for atlas creation is the acquisition of diffusion weighted MR images that are free from geometric distortions. We have integrated parallel imaging methods to collect high-resolution, near-isotropic voxel images with whole brain coverage at 3T with relatively high SNR. Parallel imaging methods reduces geometric distortions, but at 3T still require post-processing to further reduce these artifacts. As a part of this work, we have also developed and implemented an algorithm that successfully corrects this type of distortions in MR images at high magnetic fields. Average shape and parametric atlases constructed from these distortion free diffusion weighted images will be presented as well as some preliminary work on compact representations of shape and parameter (ADC and FA) variability within the population based on Active Appearance Models.
June
2005 |
Xiang Chen
Doctoral Student, Computational Biology, Carnegie Mellon University
"Robust Interpretation of Protein Subcellular Location Patterns in 3D Fluorescence Images"
Date: Friday, June 3, 2005
Time: 9:00 - 10:00 AM EST
Location: 110 Clark Hall
No podcast available
Abstract:Proteomics is the current focus of system biology and location proteomics is an important branch of it, which systematically studies protein subcellular spatial distributions for all proteins expressed in a certain cell type. Traditionally images are analyzed by visual inspection, which suffers from inefficiency and inconsistency. Automated and objective interpretation approaches are in need for location proteomics. We have designed numerical features to describe location patterns in microscope images and developed automated classifiers that distinguish major subcellular patterns with high accuracy (including patterns not distinguishable by visual examination). In addition we have described an automated method that constructs an effective partitioning of the proteins by location based on their location features. An application of these methods in molecular biology study is also covered.
