Open House 2008 Poster Session Abstracts

MSI Open House 2008 Poster Session Abstracts

Poster Session Abstracts

The poster descriptions below are presented in alphabetical order based on the name of the person who submitted the poster. Each entry includes, when provided:

  • Title of the poster,
  • Person or team presenting,
  • Organization affiliation of the presenter(s), and
  • Abstract of the presentation.


Aiman Alshare

Solar Energy Group, Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

Leucine-Rich Alpha-2-Glycoprotein-1, a Potential Serum Biomarker for Ovarian Cancer, is Secreted by an Ovarian Cancer Cell Line

John D. Andersen, B. Witthuhn, K. M. Harrington, B. Misemer, R. Jemmerson, and A.P.N. Skubitz

Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota, USA

An estimated 22,000 women will be diagnosed with ovarian cancer in the U.S. in 2008 and ~16,000 will die. Incidence and mortality statistics for ovarian cancer have changed little in the past 30 years, indicating that current methods for detection need to be improved upon. New biomarkers that either replace or are used in conjunction with the current ovarian cancer diagnostic antigen, CA125, are needed for early detection. By utilizing the mass spectrometry-based proteomic techniques of iTRAQ and Differential in-gel electrophoresis (DIGE), we identified leucine-rich alpha-2-glycoprotein-1 (LRG1) as one of the many upregulated proteins present in sera samples of ovarian cancer patients. LRG1 levels in the sera of 60 ovarian cancer patients were quantified by an ELISA and were shown to be significantly higher than the levels in the sera of noncancer women. In order to determine whether the ovarian cancer tumor itself could be the source of the LRG1 in the sera, we undertook a series of experiments. First, gene microarray analysis was performed on ovarian cancer tissues and normal ovaries; LRG1 was found to be expressed approximately 3-fold higher in the tumors. We also performed reverse-transcriptase polymerase chain reaction (RT-PCR) and quantitative-PCR using ovarian cancer tissues, normal ovaries, ovarian cancer cell lines, and normal ovarian surface epithelial cell lines. In each case, we found that the levels of mRNA were higher in the cancer samples compared to their normal counterpart. We then used these same tissues and cell lines to investigate the expression levels of LRG1 protein. By Western immunoblotting, immunocytochemistry, and fluorescence-activated cytometric sorting (FACS), we found that LRG1 protein expression was upregulated in the ovarian cancer samples. Finally, by mass spectrometry, we identified several peptides derived from LRG1 in the spent media of an ovarian cancer cell line, NIH:OVCAR5. Taken together, our results provide evidence that LRG1 protein found in the sera of ovarian cancer patients may be derived from the ovarian cancer tumor itself. Furthermore, these data suggest that LRG1 may serve as a potential biomarker in a diagnostic assay for ovarian cancer and/or as a potential target for therapeutic treatment.

Cortical Thickness Differences Between Patients With Schizophrenia and a Control Population

C.J. Bell, K.O. Lim, B.A. Mueller, R.L. Muetzel, M.L. Nelson, and T.J. White

Department of Psychiatry, University of Minnesota, Minneapolis, Minnesota, USA

Several studies have shown thinner cerebral cortex in patients with schizophrenia. The most commonly reported regions have been the prefrontal and temporal cortices. Overall however, the affected regions have been variable. This poster will examine the thickness differences in a population with schizophrenia using two methods. A General Linear Model (GLM) test of 35 bilateral cortical segmentations derived from the brain anatomical parcellation tool Freesurfer, using age as a covariate. As well as a voxel-based GLM analysis of cortical thickness using the Freesurfer tool Qdec. The results show widespread regions of thinner cerebral cortex in patients with schizophrenia compared to a control population, in agreement with some of the relevant literature.


Elhabib Benlbabib

Department of Dermatology, University of Minnesota, Minneapolis, Minnesota, USA

Patient-Specific Simulations of Bi-Leaflet Mechanical in an Anatomic Aorta

Iman Borazjani, Liang Ge, and Fotis Sotiropoulos

Department of Civil Engineering and Saint Anthony Falls Laboratory, University of Minnesota, Minneapolis, Minnesota, USA

The flow through a bi-leaflet mechanical heart valve (BMHV) implanted in an anatomic aorta is investigated by high-resolution fluid-structure interaction (FSI) simulations. The FSI solver is validated by simulating flow through a BMHV implanted in a model straight aorta similar to the parallel experimental set up at Georgia Tech Cardiovascular fluid mechanics laboratory. The numerical results are in excellent agreement with the PIV measurements. The straight and anatomic aorta are compared in terms of leaflets motion and the flow physics.

Numerical Simulations of Aquatic Locomotion

Iman Borazjani and Fotis Sotiropoulos

Department of Civil Engineering and Saint Anthony Falls Laboratory, University of Minnesota, Minneapolis, Minnesota, USA

High-resolution numerical simulations of fish-like and plankton-like swimming are reported. For fish-like swimming, some important aspects of swimming, which are not possible to investigate experimentally due to difficulties of controlling the live fish or measuring forces, are investigated through controlled numerical experiments. The hydrodynamic performance and wake structure of a mackerel and a lamprey are compared under different flow conditions. For plankton-like swimming, the hydrodynamic forces and flow field around a tethered copepod are analyzed. The 3D structure of the toroidal vortices, which were observed experimentally, are shown and the antenna was found to be the major contributor to thrust and drag among all of the copepod appendages.

Identification of Candidate Biomarkers for Ovarian Cancer Using Serum Depletion and Complementary Proteomic Techniques

John D. Andersen (1), Kristin L.M. Boylan (1), Feifei S. Xue (1), Lorraine B. Anderson (2), Bruce A. Witthuhn (2), Todd W. Markowski (3), LeeAnn Higgins (2), Yanji Xu (4), and Amy P.N. Skubitz (1)

  1. Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA
  3. Advanced Genetics Center, University of Minnesota, Minneapolis, Minnesota, USA
  4. Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA

Ovarian cancer is the fifth leading cause of cancer death for women in U.S., yet the survival rates are over 90% when ovarian cancer is detected in its earliest stage. Unfortunately, current screening techniques for ovarian cancer are neither adequately sensitive nor specific. Novel serum protein markers are needed to detect ovarian cancer in its earliest stage as well as for detection of recurrence. The discovery of new biomarkers is hindered by the presence of a small number of highly abundant proteins that comprise approximately 95% of serum total protein. In this study, three affinity columns that remove highly abundant proteins from serum were compared for their ability to reduce the dynamic range of proteins present and thereby enhance the identification of biomarkers. Low abundance serum proteins from pooled sera of ovarian cancer patients and non-cancer controls were compared using two quantitative mass spectrometry-based techniques: Differential In Gel Electrophoresis (DIGE) and isobaric tags for relative and absolute quantitation (iTRAQTM). Overall, a greater number of low abundance proteins were identified when serum samples were depleted with the ProteomeLabTM IgY 12 affinity columns than with the MARS column. Several proteins with increased levels in sera of ovarian cancer patients were identified by each method, a subset of which overlap. Current studies are underway to determine their usefulness as biomarkers for ovarian cancer.

Twisted Gastrulation Deficient Mice Phenotypic Analysis of Long Bones Using μCT Technology

Ann Carlson (1), Raj Gopalakrishnan (1), and Ravi Chityala (2)

  1. University of Minnesota, Department of Diagnostic and Biological Sciences, Minneapolis, Minnesota, USA
  2. University of Minnesota, Minnesota Supercomputing Institute, Minneapolis, Minnesota, USA

Twisted Gastrulation Protein (Twsg) is an important component of the extracellular matrix in bone. This protein has been shown in vitro to be involved in reducing osteoblast differentiation via inhibiting the signaling of bone morphogenetic proteins (BMPs). Twisted gastrulation deficient animals Twsg (-/-) were much smaller then their normal siblings and had an osteporotic phenotype when viewed by faxitron analysis. A study was undertaken to characterize the long bone phenotype and determine if there was a reduction in overall bone volume that would in part explain the osteoporotic phenotype. Femur and tibias of 3 month old sibling pairs were selected that reflected both the normal and altered Twsg state for both males and females. Micro Computed Tomography (µCT) was performed by the Physiological Imaging Center at Mayo Clinic, Rochester MN. The study revealed a significant reduction in both cortical and trabecular bone types. The Twsg (-/-) animals showed reduced trabecular volume, reduced cortical area, decreased trabecular thickness, reduced trabecular number and an increase in trabecular separation compared to Twsg (+/+). This indicates less spongy bone present in the marrow cavity and overall a very thin wall of the shaft. Studies are in process to understand the molecular pathways influencing the changes in osteoblast and osteoclast cell growth.


Pierre Carrier

Department of Chemical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

We present a generalized methodology to compute high temperature elastic constants for any crystalline structure. The elastic constants are determined based on the calculation of the materials Gibbs free energy, which can be obtained by combining first-principles total energy methodologies and the quasiharmonic approximation. A description of the algorithm for determining the elastic constants, the overall structure of the C++ program, and applications of the method for two major geophysical materials, perovskite and post-perovskite, will be described.


Monica Christiansen and Mark Wang

Department of Geology and Geophysics, University of Minnesota, Minneapolis, Minnesota, USA

VLab: A Cyberinfrastructure for Parameter Sampling Computations in Materials Science

Cesar R. S. da Silva (1), Pedro R. C. da Silveira (1), Bijaya B. Karki (2), and Renata M. Wentzcovitch (1,3)

  1. Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Computer Science, Louisiana State University, Baton Rouge, Louisiana, USA
  3. Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota, USA

We show an overall view of the service oriented architecture (SOA) used in VLab, a system aimed to handle execution of extensive workflows composed by concurrent calculations of Geo-Materials. The multiplicity of tasks composing the workflow comes from parameter sampling calculations comprising hundreds or even thousands of tasks. We review the algorithms of physical importance that underlie the system, services and metadata requirements. The system architecture emerges naturally. The SOA is a collection of web-services providing access to distributed computing nodes, controlling workflow execution, monitoring services, and providing data analyses tools, visualization services, data bases, and authentication services. A usage view diagram is described. We also describe the metadata metaphor, which can be used for a variety of other parameter sampling calculations. We also show how analysis tools, not originally developed for VLab are integrated in the SOA.

Metadata Management for Distributed First Principles Calculations in VLab: A Collaborative Grid/Portal System for Geomaterials Computations

Pedro R. C. da Silveira (1), Cesar R. S. da Silva (2), and Renata M. Wentzcovitch (1,2)

  1. Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota, USA

We describe the metadata and metadata management algorithms necessary to handle the concurrent execution of multiple tasks from a single workflow in a collaborative service oriented architecture environment. Metadata requirements are imposed by the distributed workflow that calculates elastic properties of materials at high pressures and temperatures. We explain the basic metaphor underlying the metadata management, the receipt. We also show the actual java representation of the receipt, and explain how they are XML serialized to be transferred between servers and stored in a database. We also discuss how the collaborative aspect of the user activity on running workflows could potentially lead to rush conditions, how this affects the requirements on metadata, and how these rush conditions are avoided. Finally, we describe an additional metadata structure complimentary to the receipts that contains general information about the workflow.

The Pathway Prediction System

Lynda Ellis (1), Jeff Gao (1), Kathrin Fenner (2), and Larry Wackett (3)

  1. Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Environmental Chemistry, Eawag/ETC Zurich, Dubendorf, Switzerland
  3. Department of Biochemistry, Molecular Biology and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA

The University of Minnesota Pathway Prediction System (UM-PPS) recognizes functional groups in organic compounds that are potential targets of microbial catabolic reactions, and predicts transformations of these groups based on biotransformation rules. Rules are based on the University of Minnesota Biocatalysis/Biodegradation Database and the scientific literature.

As rules were added to the UM-PPS, more were triggered at each prediction step. The resulting combinatorial explosion is being addressed in four ways: aerobic likelihood, relative reasoning, super rules, and increased rule stringency. Biodegradation experts give each rule an aerobic likelihood value of Very Likely, Likely, Neutral, Unlikely, or Very Unlikely. Users now can choose whether they view all, or only the more aerobically likely, predicted transformations. Relative reasoning, allowing triggering of some rules to inhibit triggering of others, was implemented. Rules were originally assigned to individual chemical reactions. In selected cases, these have been replaced by super rules, which include two or more contiguous reactions that form a small pathway of their own, such as beta oxidation. Some genral rules (hydrolytic dehalogenation, decarboxylation, epoxidation) were made more stringent, which improve predictions in certain cases. Rules are continually modified to improve the prediction accuracy; increasing rule stringency can improve predictions and reduce extraneous choices.

Advanced Numerical Simulations of Turbulent Flows and Transport Processes Around Surface-Mounted Obstacles

Cristian Escauriaza, Joongcheol Paik, and Fotis Sotiropoulos

Department of Civil Engineering, Minneapolis, Minnesota, USA

We perform numerical simulations of turbulent flows in complex geometries using turbulence models that can resolve large-scale vortical structures that have the largest influence on transport processes in the environment. To study the effect of these complex flows in natural rivers and streams, we develop a sediment transport model to reveal the effects of the coherent structures on the initiation of motion and bed-load transport, and incorporate an erosion model to predict the localized erosion around the obstacles. All these computational tools constitute an advanced framework to study the fundamental mechanisms of sediment transport and scour in real-life engineering problems.

Phosphorylation-Induced Structural Changes in Smooth Muscle Regulatory Light Chain

David J. E. Kast, L. Michel Espinoza-Fonseca, Andrew R. Thompson, and David D. Thomas

Dept. of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA

We are using site-directed spectroscopic labeling, EPR, fluorescence, and molecular dynamics simulations to determine the phosphorylation-induced structural transition in smooth muscle myosin regulatory light chain (RLC). Smooth muscle is activated through phosphorylation of Ser 19 on RLC, but the N-terminal 24 amino acids of RLC do not appear in any crystal structure. EPR experiments (Nelson et al., 2005) have shown that phosphorylation induces a disorder-to-order transition within the N-terminal phosphorylation domain of the RLC, in which increased helical ordering relieves inhibitory head-head interactions. To define this structural change in atomic detail, we are combining molecular dynamics simulations with spectroscopic distance constraints. Simulations on the unphosphorylated 25-residue N-terminal fragment of the RLC reveal a disordered region between residues K11-Q15, while the phosphorylated N-terminal domain maintains strong α-helicity over the same residues. The same disorder-to-order transition has been observed by both simulations extended to include the entire RLC in complex with a portion of the myosin heavy chain, as well as by dipolar EPR measurements performed on the phosphorylation domain of di-cys mutant chicken gizzard RLC. Furthermore, we have employed FRET distance measurements on di-cys mutant RLC to provide geometric constraints for the simulations. This allows both the structure and dynamics of the regulatory domain in the absence and presence of phosphorylation to be determined.

Data Mining for Connecting Genomic Data and Disease

Gang Fang, Gowtham Atluri, Rohit Gupta, Michael Steinbach, and Vipin Kumar

Department of Computer Science, University of Minnesota, Minneapolis, Minnesota, USA

One of the important potential benefits of the genetic revolution is the possibility of personalized medicine, i.e., using detailed genomic information about a person for the detection, treatment, or prevention of disease. The recent availability of individual genomic information - typically in the form of Single Nucleotide Polymorphisms (SNPs) - offers one route for making this possibility a reality. In particular, the increasing availability of SNP data has created opportunities for discovering important connections between disease and genomic factors. Although there has been some success in finding such connections with currently available techniques, these approaches have a number of limitations and are most useful for finding connections involving only one or two SNPs. This poster describes our research to develop and apply data mining techniques to find more general patterns that capture connections between SNPs and disease, including patterns that may involve a relatively large number of SNPs and patterns that show variation from patient to patient, either because of missing data or natural variation.

Regulatory Element Identification in Subsets of Transcripts: Comparison and Integration of Current Computational Methods

Danhua (Flora) Fan (1), Peter B. Bitterman (1), and Ola Larsson (2)

  1. Department of Medicine, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Biochemistry, McGill University, Montreal, Quebec, Canada

Regulatory elements in mRNA play an important role in post-transcriptional regulation of gene expression. Although there are several analytical methods that can be used to detect conserved mRNA regulatory elements in a set of transcripts, there has been no systematic study of how well any of these methods perform individually or as a group. We therefore compared how well three algorithms, each based on a different principle (enumeration, optimization or structure/sequence profiles), can identify elements in unaligned untranslated sequence regions. Two algorithms were originally designed to detect transcription factor binding sites: Weeder and BioProspector; and one designed to detect RNA elements conserved in structure: RNAprofile. Three types of elements were examined: i) elements conserved in both primary sequence and secondary structure, ii) elements conserved only in primary sequence, and iii) microRNA targets. Our results indicate that each method has advantages when used to identify a spectrum of RNA regulatory elements; and that integrating the output from all 3 algorithms leads to the most complete identification of elements. To facilitate future laboratory-based studies of co-regulation mechanisms, we developed an approach presented as a web service that can be used to identify the most promising elements from genome-wide post-transcriptional profiling data sets. This study is the first to systematically compare and integrate outputs from current algorithms for post-transcriptional mRNA regulatory element discovery, providing an approach to evaluate and incorporate new algorithms in studies of post-transcriptional control.

Multi-modality Data Mining: Accelerating Discovery

Tushar Garg (1), Ryan Muetzel (2), Michael Steinbach (1), Vipin Kumar (1), Kelvin Lim (3), Monica Luciana

  1. Department of Computer Science, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Radiology, University of Minnesota, Minneapolis, Minnesota, USA
  3. Department of Psychiatry, University of Minnesota, Minneapolis, Minnesota, USA
  4. Department of Psychology, University of Minnesota, Minneapolis, Minnesota, USA

Kelvin Lim and his team, in collaboration with Dr. Monica Luciana (Psychology) are studying normal adolescent brain development in 200 healthy subjects, age 9-23, using brain imaging, cognitive measures and genomics. Such a dataset requires extensive computational resources for image processing and new analytic strategies for identifying patterns and relationships in multi-modal data. Using the data produced by the NIDA grant and funded by by a BICB grant, Dr. Lim and his team are using expertise of Vipin Kumar's data mining group at UMN, as well as expertise at IBM, together with the computational resources of MSI, to develop new techniques for accelerating the discovery process. In this poster, we highlight results of the data mining methods developed by Dr. Kumar and his collaborators at UMN and IBM to the analysis of these multimodal datasets.

Multi-scale Modeling and Simulation of Turbulent Reacting Multiphase Flows

Michael Buhlmann, Shankhadeep Das, Saurav Mitra, and Sean C. Garrick

Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

Nanoparticles are considered one of the building blocks of nano- and micro-structured materials. The economical and efficient production of nanoparticles with prescribed/desired properties is a critical element to the implementation of nano-structured materials. Unfortunately, our predictive ability is quite limited. The difficulties are in part due to the vast range of length (and time) scales encompassing particle growth - from a few atoms to hundreds of molecules, to bulk, to integral length-scales. The work at the Computational Transport Phenomena Laboratory utilizes theory, numerical methods and experimental observations in the areas of fluid dynamics, combustion, and particle formation and growth processes to create both mathematical and physical models as well as computational tools that describe nanoparticle dynamics in a wide variety of flow-regimes, with a goal characterizing and controlling the properties and morphology of particles given thermo-chemical and thermo-physical histories. This poster will feature recent work on particle formation and growth at the nano scale and fluid-particle interactions at the micro scale.

Heat Transfer Studies in a Turbine Cascade with Inlet Skew

Kalyanjit Ghosh and Richard Goldstein

Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

The motivation for this study is to understand the effects of moving parts on the heat transfer in a gas turbine engine. A single turbine "stage" consists of a row of stationary blades, called the "stator," which "pre-turns" the flow as it spins the next row of blades called the "rotor." An actual turbine consists of a number of these "stages" to increase the total power output. This relative motion between the stator and the rotor blades greatly affects the losses in the turbine. However, most experimental heat transfer studies which have been performed to date have used a row of "stator" blades, without any moving parts, due to the inherent measurement challenges, leading to results that may not be entirely applicable to gas turbines. In this study, numerical simulations have been performed in a turbine cascade with an inlet skew. The inlet skew is imparted to the main flow by a moving wall. The Reynolds stress term is modeled using the k-omega turbulence model. Corresponding heat transfer studies show the effect of this skew on the blade surfaces and the endwall.

Exploring L1 Ligase Intrinsic Flexibility and Its Impact for Catalysis

George M. Giambasu, Tai-Sung Lee, and Darrin M. York

Department of Chemistry, University of Minnesota, Minneapolis, Minnesota, USA

Large scale molecular simulations play a distinctive role in revealing important atomic resolution details of biochemical processes, especially those that involve extremely flexible species like ribozymes. We explore L1 Ligase aptazyme intrinsic flexibility and its impact for catalysis in a biologically relevant environment through more than 500ns of large scale Molecular Dynamics simulations. These are ones of the longest simulations of an RNA system to be published up to date. L1 Ligase molecular switch is an unique ribozyme in that it specifically and regioselectively catalyze the 5' to 3' phosphodiester bond ligation, making non-canonical base pairs with its substrate. Intriguingly, the first crystal structure of an L1 Ligase construct reveals two different conformers - one thought to be active, the other inactive. This set the basis for understanding of L1 Ligase flexibility and its role in catalysis but also raised lots of questions [Robertson and Scott, Science 315 (2007) 1549]. In our simulations, the inactive conformation shows a large degree of intrinsic flexibility, showing specific contacts between two stems that are not revealed in the crystal structure. The non-canonically base-paired ligation site shows a high degree of variability observed in visiting several states. We correlate these different states with the in-line attack fitness of the ligation site.

An Exploration in Virtual Reality for the Purpose of Medical Device Design

Arthur Erdman, Eric Jerke, and Nathan Handel

Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

The objective of the project was to create 3D cardiac models from CT data using Mimics software. The 3D models were able to be exported both to a computer aided design (CAD) environment and a virtual reality environment. Furthermore, physical models were constructed using both stereolithography techniques and by creating a mold of desired cardiac regions for device testing. The molds simulated the cardiac tissue properties and allowed for implantation and evaluation of devices. Inside of the virtual environment, the user was able to navigate and create a "fly-through" of the heart. The collaboration with MSI enabled the team to use the CAVE system in Walter library to visualize the virtual representation of the heart. The work was also done in collaboration with St. Jude Medical.

Structure of the DNA Deaminase Domain of the HIV-1 Restriction Factor APOBEC3G

Kuan-Ming Chen, Elena Harjes, Phillip J. Gross, Amr Fahmy, Yongjian Lu, Keisuke Shindo, Reuben S. Harris, and Hiroshi Matsuo

Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA

The human APOBEC3G (apolipoprotein B messenger-RNA-editing enzyme, catalytic polypeptide-like 3G) protein is a single-strand DNA deaminase that inhibits the replication of human immunodeficiency virus-1 (HIV-1), other retroviruses and retrotransposons. APOBEC3G anti-viral activity is circumvented by most retroelements, such as through degradation by HIV-1 Vif. APOBEC3G is a member of a family of polynucleotide cytosine deaminases, several of which also target distinct physiological substrates. For instance, APOBEC1 edits APOB mRNA and AID deaminates antibody gene DNA. Although structures of other family members exist, none of these proteins has elicited polynucleotide cytosine deaminase or anti-viral activity. Here we report a solution structure of the human APOBEC3G catalytic domain. Five alpha-helices, including two that form the zinc-coordinating active site, are arranged over a hydrophobic platform consisting of five beta-strands. NMR DNA titration experiments, computational modelling, phylogenetic conservation and Escherichia coli-based activity assays combine to suggest a DNA-binding model in which a brim of positively charged residues positions the target cytosine for catalysis. The structure of the APOBEC3G catalytic domain will help us to understand functions of other family members and interactions that occur with pathogenic proteins such as HIV-1 Vif.

Visualizing Computational Number Theory

Ben Jordan

School of Mathematics, University of Minnesota, Minneapolis, Minnesota, USA

Number theoretic algorithms are ubiquitous in computing, due not only to their role in cryptography, but to the central role that divisibility plays in mathematics. I am investigating methods for visualization and analysis of the structure of discrete, countable sets, such as the integers (Z), the rationals (Q), the Gaussian integers (J), and some relevant subsets. I discuss and demonstrate computational methods used to find symmetry, reducibility, bulk probability, and self-similarity in these sets. Applicability in factoring problems is discussed, as well.

Abnormal Compressibility of Ferropericlase Across Its Spin Transition

João Francisco Justo (1), Zhongqing Wu (1), Cesar R. S. da Silva (1), Dave Yuen (2), and Renata Wenztcovitch (1)

  1. Department of Chemical Engineering and Materials Science, Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Geology and Geophysics, Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA

The thermoelastic properties of ferropericlase Mg1-xFexO (x = 0.1875) across the iron high-to-low spin crossover have been investigated by first principles at Earth's lower mantle conditions. The transition has important consequences for the elasticity such as a substantial bulk modulus reduction. At room temperature the transition is quite sharp in pressure but broadens with increasing temperature. Along a typical geotherm the transition should occur across most of the lower mantle with a more significant bulk modulus reduction (of about 25%) around 1400-1600 km depth. This softening would also cause a large reduction in the activation free energy of the ferropericlase rheology, leading to a viscosity minimum in the mid-lower mantle, in agreement with results from geoid inversion and postglacial rebound studies.

Rethinking Relativistic Particle Energization in the Inner Magnetosphere: A Study of Nonlinear Electron Acceleration and Scattering in the Presence of Large-Amplitude Whistler Waves

Kris Kersten, C. Cattell, J. Wygant, K. Goetz, and I. Roth

Department of Physics, University of Minnesota, Minneapolis, Minnesota, USA

Recent waveform captures from the University of Minnesota S/WAVES instrument on the STEREO-B satellite have revealed large-amplitude whistler-mode waves in the Earth's radiation belts, prompting a reinvestigation of particle energization mechanisms. Previous studies assume small amplitude waves that impart energy to radiation belt electrons through a chain of stochastic interactions. As such, the existence of relativistic electrons in the radiation belts was thought to be due to multiple wave-particle interactions over the course of hours or even days. Using adapted particle tracing codes originally designed to study resonant enhancements of energetic electron fluxes we have modeled nonlinear wave-particle interactions under conditions seen by the STEREO satellites, showing that electrons can be accelerated by several MeV in just a few tenths of a second. This demonstration of rapid particle energization suggests that current quasi-linear models do not accurately describe the dynamics of the Earth's radiation belts. The simulations have also showed evidence of rapid pitch angle scattering potentially leading to particle precipitation and loss in the ionosphere, consistent with low altitude SAMPEX observations of short-lived enhancements in the precipitation of energetic electrons. In addition to a brief summary of the STEREO S/WAVES observations that motivate this investigation, we will describe the basics of the particle tracing code and present results of the simulation across a range of plasma and particle parameters. We will also outline current work on the implementation of a more realistic simulation model and a study of the impact of large-amplitude waves on particle distributions. Ultimately, this study will provide information important to the design of specific instrument modes on the University of Minnesota/U.C. Berkeley Electric Field and Waves (EFW) instrument on NASA's Radiation Belt Storm Probes (RBSP) spacecraft to be launched spring of 2011.

On The Nature of Antimicrobial Activity: A Model for Protegrin-1

Allison A. Langham, Abdallah Sayyed Ahmad, Yiannis N. Kaznessis

Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota, USA

Though antimicrobial peptides have been studied for more than two decades as possible substitutes for traditional antibiotic drugs, their mechanism of action is still not fully understood. We have performed over 150ns of simulation of a protegrin-1 (PG-1) pore in a lipid bilayer composed of plamitoyloleoylphosphatidylethanolamine (POPE) and palmitoyloleoylphosphatidylglycerol (POPG) lipids meant to mimic the inner membrane of a bacterial cell. The simulations improve on a model of an octomeric pore proposed from NMR experiments. From the results we determine that the pore more closely follows the barrel-stave model than the toroidal model for insertion into the bilayer. We explore the movement of ions through the pore in detail. The pore allows negatively charged chloride ions to pass through at an average rate of one ion per two nanoseconds. We observe only two events of sodium ions crossing through the pore. Potential of mean for was calculated for the water and rough the pore as easily as the water molecules. We explore the potential for PG-1 to kill cells through osmotic lysis or destabilization of transmembrane potential.

Origin of Mutational Effects At Various Positions on Hammerhead Ribozyme Catalysis From Molecular Dynamics Simulations

Tai-Sung Lee, George M. Giambasu, and Darrin M. York

Department of Chemistry, University of Minnesota, Minneapolis, Minnesota, USA

A series of twenty-four 100 ns (total: 2.4 micro seconds) molecular dynamics (MD) simulations of the native and mutated full length hammerhead ribozymes in the reactant state and in an activated precursor state (G8:2'OH deprotonated) are reported. Mutant simulations at the C3, G8, and G5 positions are performed. The results support a mechanism where the 2'OH of G8 acts as a general acid catalyst and suggest that the origin of the observed mutation effects at these positions is the distortion of the Watson-Crick hydrogen bonding between G8 and C3, or the distortion of the 3 key hydrogen bonds holding the relative positions between C17 and G5.

Evaluation of the Transition State Structure in the Enzymatic Farnesylation Reaction

Stepan Lenevich, Christopher Cramer, and Mark D. Distefano

Department of Chemistry, University of Minnesota, Minneapolis, Minnesota, USA

Protein farnesyltransferase (PFTase) catalyzes the attachment of farnesyl moiety to a specific cysteine in a number of proteins. Oncogenic Ras proteins require the farnesylation of a cysteine group to be functional and hence PFTase inhibitors can potentially be used as anticancer therapeutics. Since enzymes significantly stabilize the transition state and transition state analogs can form particularly tight complexes with enzyme, they are attractive candidates for anticancer drugs development.

Currently we are employing kinetic isotope effect (KIE) measurements to evaluate the mechanism and the structure of the transition state in the enzymatic farnesylation reaction. The KIEs in model and enzymatic reactions are measured employing different methods. Then the KIEs are computed using Gaussian 03 program. The transition state structure of the geranylation reaction is optimized, the key bond lengths in the TS structure are adjusted and the KIEs for new structures were calculated. The structure with calculated KIEs close to the experimentally measured ones resembles the transition state in enzymatic farnesylation reaction. This approach allows us to calculate the electron density map of the transition state of the enzymatic farnesylation reaction and provides insight for the design of putative anticancer drugs.

Historical Census Record Linkage Using Support Vector Machines

Ron Goeken, Tom Lenius, and Rebecca Vick

Minnesota Population Center, University of Minnesota, Minneapolis, Minnesota, USA

Record linkage entails identifying the same individuals in different data sets. Using a customized version of the Freely Extensible Biomedical Record Linkage (Febrl) software, distance scores are generated for pairs of values representing first name, last name, and age. Subsequently, a Support Vector Machine (SVM) processes the score values to identify links. The SVM is trained by providing score data containing examples of links and non-links. The SVM then can be used to classify links in the test data.

Brownian Dynamics Simulations of Tethered Polymers on Curved Surfaces

Margaret Linak, Martin Kenward, and Kevin D. Dorfman

Department of Chemical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

Surface tethered polymers are an important component in many physical systems, including coating applications, microfluidic devices, drug delivery vehicles, and molecular targets in DNA microarrays. We present a study of an isolated polymer chain tethered to a curved, impenetrable surface, where the radius of gyration is varied from highly concave, through flat, to highly convex. Utilizing Brownian dynamics simulation, we examine the equilibrium properties of the polymer as a function of its molecular size and degree of curvature.

Biomechanics of Erections

Ahmed Mohamed (1), Gerald Timm (2), and Arthur Erdman (1)

  1. Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Urological Surgery, University of Minnesota, Minneapolis, Minnesota, USA

Most models of the erect penis designed the shaft as a one-compartment thin walled pressurized vessel fixed at a one point, when in reality it is a two-compartment thin walled pressurized vessel in which the compartments diverge as they enter the body and are fixed at two separate points. But, there isn't a clear understanding as to why Mother Nature chose the latter. To gain a better understanding of the biomechanics of the penis and its structural integrity, we designed and developed refined two-dimensional and three-dimensional models of the erect penis using Finite Element Analysis (FEA) with varying anatomical considerations for analyzing structural stresses and buckling. Preliminary results have indicated that under axial and lateral loading, an erection supported by a two-compartment thin walled pressurized vessel was more structurally stable than the one-compartment vessel. We will validate our results by building physical models of our computer simulated models and then measure the buckling and rigidity of each physical model. This project will be first of its kind to model the penis with two corpora cavernosal bodies fixed at the pelvic arch in order to achieve the most anatomically accurate model of an erect penis.

Relationships Between Brain Structure and Bimanual Task Performance in Healthy Adolescents

Ryan Muetzel (1), Paul Collins (2), Kelvin Lim (3), and Monica Luciana (2)

  1. Department of Radiology, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Psychology, University of Minnesota, Minneapolis, Minnesota, USA
  3. Department of Psychiatry, University of Minnesota, Minneapolis, Minnesota, USA

The structural brain changes associated with adolescent development have been studied primarily with conventional magnetic resonance imaging (MRI) and to a lesser extent, diffusion tensor imaging. With the exception of functional imaging, limited work has examined relationships between brain development and behavioral change. The current study incorporated a longitudinal design to examine the development of bimanual motor performance in relation to brain structure. Healthy adolescents (n=64) were scanned twice, with a two-year period between scans. Bimanual performance was assessed at both time-points. Automated tissue parcellation (FreeSurfer) was applied to T1-weighted images yielding gray and white matter volume for pre- and post-central gyri. Volumes for three regions of the corpus callosum were also computed. Changes in tissue over the two-year time span were assessed by computing difference scores for each region between sessions. The same approach was used for bimanual motor performance. Significant relationships were observed between bimanual task performance and white matter changes in the pre- and post-central gyri. These preliminary findings relate structural brain characteristics to behavioral performance using a longitudinal design.

Computational Approaches for Protein Function Prediction

Gaurav Pandey (1), Michael Steinbach (1), Rohit Gupta (1), Tushar Garg (1), Vipin Kumar (1), Bhupinder Juneja (1), Scott Fahrenkrug (2), and Yanji Xu (3)

  1. Department of Computer Science, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Animal Science, University of Minnesota, Minneapolis, Minnesota, USA
  3. Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA

Proteins are the most essential and versatile macromolecules of life, and the knowledge of their functions is a crucial link in the development of new drugs, better crops, and even the development of synthetic biochemicals such as biofuels. Experimental procedures for protein function prediction are inherently low throughput and are thus unable to annotate a non-trivial fraction of proteins that are becoming available due to rapid advances in genome sequencing technology. This has motivated the development of computational techniques that utilize a variety of high-throughput experimental data for protein function prediction, such as protein and genome sequences, gene expression data, protein interaction networks and phylogenetic profiles. In this poster, we discuss our work on several problems related to the computational prediction of protein function.

Simulation of RCS-Aerodynamic Interaction of Mars Science Laboratory Capsule

David M. Peterson and Graham V. Candler

Department of Aerospace Engineering and Mechanics, University of Minnesota, Minneapolis, Minnesota, USA

Numerical simulations are performed of the Mars Science Laboratory (MSL) capsule as it enters the atmosphere of Mars at 18 times the speed of sound. The simulations are three dimensional and unsteady, using the detached-eddy simulation method to model the effects of turbulence. The goal of the work is to numerically investigate any aerodynamic interactions that occur during the firing of the reaction-control system (RCS) thrusters used to maneuver the capsule. Any such interactions could reduce the effectiveness of the RCS thrusters or otherwise alter their performance in undesirable ways. Interactions that reduce the effectiveness of the thrusters by more than 10% are found during two of the four maneuvers simulated.

Protocatechuate 3,4-Dioxygenase: In Search of Catalytic Intermediates

Vincent M. Purpero and John D. Lipscomb

Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA

Protocatechuate 3,4-dioxygenase (34PCD, E.C. is a non-heme Fe3+ dependent enzyme involved in aromatic metabolism. The substrate of 34PCD is protocatechuic acid (PCA, 3,4-dihydroxybenzoic acid), which is a central metabolite in the β-ketoadipate pathway for aromatic degradation. 34PCD is an intradiol dioxygenase that incorporates both atoms of molecular oxygen into the product (β-carboxy-cis,cis-muconic acid), while cleaving the aromatic ring of PCA.

Mutations of inner sphere ligands significantly slow catalysis, however each mutant affects turnover through different mechanisms. Tyrosine 447 is proposed to act as a base in deprotonation of the incoming substrate as well as protonation of the product. In addition, tyrosine 447 dissociates from the metal ion during catalysis, whereas the crystal structure of the enzyme-substrate complex demonstrates that the mutation to a histidine is 2Å shorter and does not ligand with the metal center. Crystallization in aerobic conditions of Y447H and the substrate analogue 4-amino-3-hydroxybenzoate, shows substrate in two active sites and what is proposed to be a product in the thrid active site. This is the first time it has been shown with 34PCD, that multiple species can exist in crystallo in different active sites.

Alternatively, tyrosine 408 is trans to a hydroxyl group, which upon mutating to a histidine changes the strength of that bond and is proposed to alter the trans effect on the metal center. Crystallography has shown the tyrosine 447 bond becomes shorter due to this mutation at residue 408 and this can hinder the bidentate chelation of the substrate.

The Y447H/H462Y double mutant slows catalysis by 7000 fold and we now have the 2.0 A resolution crystal structure, and the chelated substrate complex. There is evidence with this enzyme and turnover with the native substrate PCA, that we can achieve the enzyme-product complex. Similarly to the Y447H-4A3H crystal, it appears there are unique species in each active site, with product in two active sites, and substrate in the other.

Solute Transfer by Advective Flows into the Sediment Beds of Streams and Lakes

Qin Qian, Vaughan R. Voller and Heinz G. Stefan

Department of Civil Engineering and Saint Anthony Falls Laboratory, University of Minnesota, Minneapolis, Minnesota, USA

Solute exchange between surface water and the sub-aqueous sediment is a fundamental hydrologic process that can affect water quality and ecology in freshwater systems. The pressure distribution at the water/sediment interface, due to either surface wave motion or bed forms, can induce a flow field in the sediment and add an advection term to the transfer process, thereby, enhancing the solute exchange between water and the sediment significantly. To quantify this enhancement, an unsteady two-dimensional (2-D) advection dispersion numerical model was developed to predict the transient concentration evolution in an initially clean bed below a contaminated water body with a constant solute composition. The development of this tool required the careful construction of novel numerical techniques to mitigate effects associated with advection-dispersion approximations. From the predicted, time-variable 2-D concentration field, one-dimensional (1-D) laterally averaged concentration profiles were calculated. These resulting 1-D concentration profiles were matched to the solution of an unsteady 1-D vertical dispersion equation. A depth-variable "enhanced vertical dispersion coefficient" (DE(y)) was thus determined by inverse modeling. The process of calculating a 1-D enhanced dispersion coefficient was investigated for two important environmental solute transport cases (1) hyporheic flow and underflow in a gravel stream bed under a standing pressure wave, and (2) in a lake bed under a progressive pressure wave. In both cases the utility of calculating DE(y) was underscored by the fact that normalizations could be found that allow the development of closed form expressions that fit the simulated values of DE(y). The expression for DE(y) in both cases were also applied in a 1-D solute exchange model and provided a very close match to available experimental data. The enhanced dispersion coefficient (DE(y)) is 10 to 1000 times larger than the hydrodynamic dispersion or molecular diffusion in a sediment bed. Knowledge of DE(y) makes it possible to analyze and/or model the transfer of conservative as well as non-conservative solutes between surface water and sediment without explicit analysis of the flow through the pore space. This is a distinct advantage for water quality modeling in surface water and pore water.

Modeling Nanodusty Plasmas

Lavanya Ravi and Steven L. Girshick

Department of Mechanical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

Plasmas are used to synthesize nanoparticles used for various applications such as photovoltaics, microelectronics. In spite of the vast literature on the topic, the early stages of nanoparticle formation and growth have not been fully understood owing to the very small sizes and timescales at which the phenomena happen. This numerical model simulates the nanoparticle growth and transport at the early stages of nanodusty plasmas.

Microscopy and Imaging Collaborations With the CBS Imaging Center and MSI

Tracy E. Anderson (1), Brian P. Piasecki (2), and Mark A. Sanders (1)

  1. CBS Imaging Center, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Genetics, Cell Biology and Development, University of Minnesota, Minneapolis, Minnesota, USA

The College of Biological Sciences Imaging Center is a fee for service facility specializing in light and electron microscopy located in Snyder Hall on the St. Paul Campus of the University of Minnesota. The Imaging Center is a member of a consortium of service centers known as Biodale offering state-of-the-art instrumentation and user-friendly, service and training. These services are available to researchers both at the University and the private sector. Biodale is located in the lower levels of Snyder Hall and Gortner Laboratory of Biochemistry, 1475-1479 Gortner Avenue, on the University's St. Paul campus.

The Imaging Center is equipped and staffed to support teaching, training and research activities. The Imaging Center specializes in light and electron optical methods with expertise centered on digital imaging. The facility focuses on providing a complete solution from sample preparation to quantitative morphological analysis of fixed and living cells using microscopy.

We will present a brief description of the equipment and capabilities we currently have in the Imaging Center. You can also visit our web site for equipment scheduling and more information.

Much of the data collected from light microscopes is intrinsically hampered by out-of-focus information and deconvolution computational techniques offer a solution to this problem. Parallel computing available through MSI allows near real-time image analysis and reconstruction of batch files. The researchers have implemented constrained iterative deconvolution methods to remove out-of-focus information and reveal more structural and functional details. The amount of processing time was substantially less compared to serial deconvolution of the data sets using local computing. Data from a recent publication (2008) by Piasecki, et. al. entitled The Uni2 phosphoprotein is a cell cycle-regulated component of the basal body maturation pathway in Chlamydomonas reinhardtii will be shown as an example of collaborative results.

Modeling the Effect of Dowel Misalignment in Concrete Pavements

Priyam Saxena and Kyle Hoegh

Department of Civil Engineering, University of Minnesota, Minneapolis, Minnesota, USA

Optimizing design and construction guidelines for Portland cement concrete (PCC) pavements is an important task in providing efficient transportation in the United States. The NCHRP 10-69 study involves development of guidelines for transverse joints in concrete pavement construction, and more specifically for dowel bar alignment tolerances. While these tolerances have not been based on pavement performance in the past, this study focuses on the effect of dowel alignment on the joint performance.

In accordance with the 10-69 project, several 3D ABAQUS models were developed to provide detailed modeling of dowel/PCC interaction. The model predictions are based on dowel/PCC interface input characteristics. In this study, the ABAQUS 3D model was calibrated and validated using the results of the modified pullout and shear pull testing of different dowel alignments tested in the laboratory.

The calibrated ABAQUS beam-dowel model was then used to simulate a virtual test for misalignment of the dowel in the actual pavement. The misalignment levels tested in the lab, and simulated by ABAQUS include different embedment lengths and concrete covers due to vertical and horizontal translations, dowel rotation in the form of vertical tilt, variable dowel diameters, as well as combinations of the misalignments. The model was run under load control and displacement control modes and compared with the laboratory results. The calibrated and verified model was then used to predict the effect of dowel misalignment for extended and full scale cases that were not able to be tested in the laboratory. The results of the ABAQUS model will be displayed and discussed on the poster presentation.

Assessment of Density Functionals, Semiempirical Methods, and SCC-DFTB for Protonated Creatinine Geometries

Nicole M. Settergren (1), Philippe Buhlmann (1), and Elizabeth A. Amin (2)

  1. Department of Chemistry, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Medicinal Chemistry, University of Minnesota, Minneapolis, Minnesota, USA

Creatinine concentrations in blood and urine can be used to detect renal insuffuciencies and muscle diseases. Our research team seeks to fabricate an ion-selective electrode (ISE) to measure creatinine with high selectivity and robustness, which current chemical sensors lack. To make such an ISE, a receptor that selectively binds protonated creatinine (creatininium) is required. Computational methods have proven valuable for the rational design of similar receptors, but systematic studies, which have not yet been reported, are necessary to understand the reliability of specific in silico methods for creatininium. Here we examine the accuracy of DFT (density functional theory) and WFT (wavefunction theory) calculations for the creatininium ion structure, validated against two experimental crystal structures. We tested twenty one local and nonlocal density functionals, Hartree-Fock theory, four semiempirical molecular orbital methods of the neglect of differential overlap (NDO) type, and two tight-binding methods. We specify the best methods for protonated creatinine and discuss the importance of hydrogen bonding elucidated by our calculations.

Visualization of Tsunami Waves With the Amira Package

Erik O.D. Sevre, Yingchun Liu, and David A. Yuen

Department of Geology and Geophysics, Minnesota Supercomputing Institute, Laboratory of Computational Science and Engineering, University of Minnesota, Minneapolis, Minnesota, USA

In recent years numerical investigations of tsunami wave propagation have been spurred by the magnitude 9.3 earthquake along the Andaman-Sumatra fault in December, 2004. Visualization of tsunami waves being modeled can yield a better physical understanding about the manner of wave propagation over realistic seafloor bathymetries. We will review the basic physics of tsunami wave propagation and illustrate how these waves can be visualized with the Amira visualization package. For modeling tsunamis we have employed both the linear and nonlinear versions of the shallow-water wave (SWW) equation. We will give various examples, illustrating how the time-history files can be loaded up by Amira to make movies, how the wave-heights of the tsunami waves can be portrayed and viewed with illumination from light sources and how movies can be used to facilitate physical understanding and give important information in the initial stages of wave generation from interaction with the ambient geological surroundings. We will show examples of tsunami waves being modeled in the South China Sea, Yellow Sea and southwest Pacific Ocean near the Solomon Islands. The last situation is important, as this will illustrate the tremendous influence exerted by the immediate geological environments close to the earthquake source in preventing waves to propagate in certain directions, in this case, toward the Hawaiian Isles.

Visualization should be a part of any training program for teaching the public about the danger of tsunami waves. We propose that interactive visualization with a web-portal would be useful for understanding more complex tsunami wave behavior from solving the 3-D Navier-Stokes equations in the near field. We will give a demo of Amira showing tsunami wave propagation on the PowerWall at the LCSE.

Excited State Dynamics in 7-azaindole: Resonance Raman Spectroscopy and Computational Simulation

Nathan R. Erickson, Molly B. Beernink, Nathaniel K. Swenson, and Jonathan M. Smith

Department of Chemistry, Gustavus Adolphus College, Saint Peter, Minnesota, USA

Excited state proton transfer dynamics in 7-azaindole (7AI) and 7AI's dimer is of interest due to its biological significance as a model system for DNA base pair interaction. This phenomena has been studied by a range of methodologies in solution and the gas phase. In our computational and experimental work using Resonance Raman spectroscopy we are afforded unique access to the early proton transfer dynamics mediated by solvent. We will present resonance Raman spectral simulations of monomer and dimer species and correlate these simulations with our isotopic spectral data. We will report the relevance of a range of computational theory and experimental results to ongoing consideration of the mechanism of the excited state proton transfer in 7AI and related systems as a function of solvent.

Phase Diagram Visualization via Continuously-Fed Crystallization: Experiments and Model

Masano Sugiyama (1) and Victor Barocas (2)

  1. Department of Chemical Engineering, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota, USA

X-ray diffraction is the most common way to determine structural information of proteins at an atomic level. Elucidation of these structures has a dynamic impact on biotechnology and pharmacology. To determine the protein structure, a high quality crystal of sufficient size is required. However, the production of such crystals is a current bottleneck. Current technology involves hundreds of conditions in this multi-parametric process to be tested by either batch or vapor diffusion crystallization techniques.

A continuous-feed crystallization chamber is a microfluidic system that allows the experimenter to screen a large range of salt and protein concentrations. This microfluidic application allows the experimenter to predict the phase diagram of a protein, which typically requires hundreds of experiments, from one experiment. A continuous-feed crystallization chamber has been successfully fabricated and characterized in terms of its flow profile. This device has successfully predicted the phase diagram for lysozyme.

Structural Dynamics of the Sarcoplasmic Reticulum Ca2+-ATPase (SERCA) Studied by FRET and Molecular Modeling

Bengt Svensson, Deborah L. Winters, Joseph M. Autry, L. Michel Espinoza-Fonseca and David D. Thomas

Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, USA

The Ca2+-ATPase (SERCA) is an integral membrane protein that uses energy from ATP hydrolysis to pump calcium into the sarcoplasmic reticulum, which relaxes the heart muscle and provides the calcium gradient needed for the next muscle contraction. Our laboratory is engaged in spectroscopic analysis of structural dynamics of the proteins involved in Ca2+ transport. The X-ray crystal structures suggest that the nucleotide-binding (N) and actuator (A) domains of SERCA move apart by 27Å upon Ca2+ binding. In order to test this hypothesis, a fusion protein, containing Cyan Fluorescent Protein (CFP) linked to the N-terminus of SERCA was constructed. Fluorescence Resonance Energy Transfer (FRET) was used to monitor the A to N interdomain distance for CFP-SERCA selectively labeled at position 515 with fluorescein isothiocyanate. We have also performed simulations of CFP-SERCA, in order to interpret FRET data obtained from this construct. To obtain a more complete view of the nature of the domain motions, we have also performed molecular dynamics simulations of the calcium-bound (E1.Ca) SERCA in an explicit water and lipid environment. The molecular dynamics simulations reveal domain motions in the protein, particularly in the nucleotide-binding and the actuator domains. We propose from our experiments presented here that, although crystal structures of SERCA are available, they represent only a single snapshot of the broad conformational landscape that the protein adopts.

Anterograde Microtubule Transport Drives Microtubule Bending in LLC-PK1 Epithelial Cells

Erkan Tuzel (1), Andrew D. Bicek (2), Aleksey Demtchouk (2), Maruti Uppalapati (3), William O. Hancock (3), Daniel M. Kroll (4), and David J. Odde (2)

  1. Institute for Mathematics and Its Applications, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota, USA
  3. Department of Bioengineering, Pennsylvania State University, University Park, Pennsylvania, USA
  4. Department of Physics, North Dakota State University, Fargo, North Dakota, USA

Microtubules are proposed to act mechanically as compressive struts that resist both actomyosin forces and their own polymerization forces to mechanically stabilize cell shape. However, individual microtubules in living cells are often highly bent, suggesting that their resistance to bending is weak. The origin of these deformations in living cells is still unknown. Recent experiments on LLC-PK1 epithelial cells strongly suggest that F-actin dynamics and polymerization of microtubules play a minor role, and that the dominant mechanism is anterograde transport via molecular motors. Quantitative analysis of these deformations using curvature distributions exhibit a characteristic exponential decay in the tail of these distributions. Interestingly, the curvature distribution of microtubules in an in vitro kinesinmicrotubule gliding assay, a system in which microtubules glide over a bed of molecular motors attached to a substrate, recapitulates the distribution of curvatures measured in vivo. Motivated by these experiments, both in vivo and in vitro, we have modeled the deformation of microtubules under the influence of molecular motor forces by coarse-grained simulations. In the simulations, microtubules are modeled as semi-flexible polymers with rigid bond constraints embedded in a Langevin heat bath. Molecular motors exert forces on the microtubules, and they walk along microtubule tracks according to their known force-velocity relations, bind and unbind stochastically. Preliminary results show that the distributions of curvatures for high density of molecular motors exhibit an exponential tail. Simulation results support the idea that molecular motors are the dominant source of bending in living cells and that molecular motors appear to generate most of the strain energy stored in the microtubule lattice.

MgSiO3 at Conditions of the Giant's Cores and Exoplanets

Koichiro Umemoto (1), Renata M. Wentzcovitch (1) and Philip B. Allen (2)

  1. Department of Chemical Engineering and Materials Science, Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Physics and Astronomy, Stony Brook University, Stony Brook, New York, USA

CaIrO3-type MgSiO3 is the planet-forming silicate, stable at pressures and temperatures (PTs) at and beyond those of Earth's core-mantle boundary (CMB). We have found using first principles quasiharmonic free energy computations that this mineral should undergo a phase transformation at PTs expected to occur in the cores of the gas giants. The transformation should also affect the thermal-chemical structure of the massive dense core of the Saturn-like exoplanet HD149026 and perhaps the mantle of Earth-like GJ876. At ~10 Mbar and ~10,000 K the resulting aggregate should have thermally activated carriers, causing electrical conductivity close to metallic values, and a correspondingly large electronic contribution to thermal conductivity, while simultaneously shutting down radiative heat transport because of electronic damping.

Molecular Dynamics Studies of the Nicotinic Acetylcholine Receptor

Hai-Long Wang and Steven M. Sine

Department of Physiology, Mayo Clinic College of Medicine, Rochester, Minnesota, USA

Communication between a cell and its environment relies on channel-forming proteins to provide a low energy pathway for ions to move in and out. Although channel-forming proteins are essential to all life forms, the atomic-scale mechanisms that enable ions to pass through the channel remain elusive due to the lack of experimental approaches to monitor the protein and ion in real time and at atomic resolution. A powerful alternative approach is molecular dynamics (MD) simulation based on the laws of physics applied to the increasing body of protein structures resolved at atomic resolution. Here we present all-atom MD simulations applied to the nicotinic acetylcholine receptor (nAChR) that initiates voluntary movement in skeletal muscle. By focusing on individual permeant cations, we find that selective cation translocation occurs in stages: cations are first selected through a series of oppositely charged residues within the protein vestibule leading to a narrow hydrophobic constriction, but then hydration of the narrow region and dynamic fluctuations of the protein enable the cation to pass through. The findings provide a general framework for understanding how ions are selected for transport based on charge, and how the dynamic interplay between water, the ion and the channel protein enable rapid ion translocation through the broad class of channel-forming proteins with hydrophobic barriers.

Effect Of Thallium Dopants on the Electronic Band Structure and Optical Properties of Csi 110 and 100 Surfaces

Victor Weir and E. Russell Ritenour

Medical Physics Division, Department of Radiology, University of Minnesota, Minneapolis, USA

The electronic band structure and optical properties of CsI 110 and 100 surfaces are calculated and analyzed using Dmol3. The band structures and are used to calculate the effective electron mass, plasma frequency and the radiant quantum efficiency. These numbers are in agreement with the known experimental results. The character of the conduction and valence bands are analyzed in terms of the types of orbitals in those regions. The changing character of these bands is in line with the expected behavior of CsI in terms of the light output and RQE.

Modeling of Prolyl-Leucyl-Glycinamide (PLG) Analogs That Modulate the Dopamine D2 Receptor

Richard L. Wood (1), Brendan J. Young-Dixon (2), Rodney L. Johnson (1) and Elizabeth A. Amin (1)

  1. Department of Medicinal Chemistry, University of Minnesota, Minneapolis, Minnesota, USA
  2. Minnesota Supercomputing Institute (MSI) summer undergraduate intern, Northwestern University, Evanston, Illinois, USA

Prolyl-leucyl-glycinamide (PLG) is a unique endogenous peptide that modulates dopamine receptor subtypes of the D2 receptor family within the CNS. We seek to elucidate the structural basis and molecular mechanism by which PLG modulates dopamine receptors, toward the development of new drugs to treat Parkinson's and related diseases of the CNS. Toward this goal, we evaluate the suitability of a wide variety of molecular mechanics (MM) force fields, semiempirical neglect of differential overlap (NDO) methods, self-consistent-charge density-functional tight-binding (SCC-DFTB) methods, density functional theory (DFT) levels, and Hartree-Fock theory for a family of novel PLG analogs designed in our laboratory, using crystal structures as benchmarks. We also present benchmark databases, obtained by coupled-cluster calculations with single, double and quasiperturbative triple (CCSD(T)) excitations, of bond distances and partial charges for a representative fragment common to our PLG analogs, and use these to test a selection of popular density functionals. We specify the best methods for this class of PLG analogs, and we recommend the M05-2X hybrid meta GGA functional to obtain accurate geometric parameters on these and similar peptidomimetic compounds.

Optimal Transvenous Coil Position for Active-Can Single-Coil ICD Defibrillation Efficacy: A Simulation Study

Fei Yang and Robert P. Patterson

The Bakken Medical Instrumentation and Device (MIND) Laboratory, Institute for Engineering in Medicine, University of Minnesota, Minneapolis, Minnesota, USA

The implantable cardioverter defibrillator with an active can and a single coil lead is effective in treating ventricular fibrillation, but the lead placement associated with high fibrillation efficacy is still controversial. In this study, an anatomically realistic finite difference model of the thorax with 3.8 million elements was developed based on MRI images to determine the effect of transvenous lead placement on defibrillation efficacy. Four electrode configurations were simulated and their defibrillation efficacy were evaluated based on a set of metrics including voltage defibrillation threshold, current defibrillation threshold, interelectrode impedance, potential gradient distribution uniformity, current density distribution and myocardium damage. It was found that the optimal electrode configuration is to position leads in the middle of the right ventricular (RV) cavity.

First Principles Study of Several Phase Transitions in Earth Transition Zone

Yonggang G. Yu (1), Zhongqing Wu (2), Renata M. Wentzcovitch (2)

  1. Department of Chemistry, Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA
  2. Department of Chemical Engineering and Materials Science, Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota, USA

First principles quasiharmonic free energy calculations have been used to study several important phase transitions in Mg2SiO4 and MgSiO3 system under Earth upper mantle conditions (P ≤ 23 GPa and T ≤ 1900 K). Encouraging agreement with experiments has been achieved in both thermodynamic properties of high-pressure single crystalline phases and relevant phase boundaries between them. Detailed comparison with experiments is given for the a - b - g transitions in Mg2SiO4, the g spinel dissociation into perovskite and periclase, and for the low-pressure to high-pressure clinoenstatite transition. Common trends displayed by these calculations are: (1) GGA phase boundaries are closer to experimentally measured boundaries than LDA boundaries; (2) LDA calculated thermodynamic properties agree much better with experiments than GGA derived properties. In overall these calculations can supplement high-experimental data on Earth mantle minerals.


Nan Zhang

Department of Urology, University of Minnesota, Minneapolis, Minnesota, USA

The main focus is a realtime surgical simulation framework. In this framework, we use two levels of simulation hierarchies: one coarse mesh for the deformation and collision detection, another fine mesh for the visual display. We use a truss model to approximate FEM simulation results while preserving the merit of fast speed of mass-spring systems. We then use a GPU-based deformation transfer algorithm to efficiently map the deformation from the coarse mesh to the fine mesh. The transfer algorithm not only transfers the deformation, but also provides the rotation matrix for rendering attached volume texture. The overall result is a fast, realistic deformation system. We are building a kidney removal procedure simulator based on the proposed framework.