|
his summer, fifteen undergraduate student researchers from across the country served ten-week internship appointments at the Supercomputing Institute, including three students in the Computational Neuroscience Program. These students were selected from a pool of 105 applicants. The students worked closely with faculty advisors on many projects.
 |
| Charles Park (left) works with Ian Tregellis (right), a graduate student in Professor Thomas Jones' Astronomy research group.
|
 |
| Coordinator of the Computational Neuroscience Internship Program, Kathleen Clinton (center), speaks with the Computational Neuroscience interns Aaron Miller (left) and David Liebelt (right).
|
The Supercomputing Institute Summer Internship Program, currently in its ninth year, promotes undergraduate involvement in ongoing and new research in scientific computing, digital technology, and visualization in the physical, medical, and social sciences and engineering and in new software development efforts for scientific computing and graphics support for such research with the main goal being to carry out useful and interesting research. This program provides an opportunity for a challenging and enriching educational experience for undergraduate students interested in pursuing graduate or professional education and research in scientific computing and/or graphics.
During the summer, interns participated in Institute sponsored tutorials specific to high-performance computing. To conclude the summer, the interns presented talks open to the entire research community. These talks allowed them to share their work and to gain experience making scientific presentations. The program allowed the students to perform research in close collaboration with faculty investigators and their research groups and to discuss research with faculty members, post-doctoral associates, graduate students, and other interns with similar interests.
Project Descriptions
Sarah Alfano, a Physics major at Kent State University, worked with Professor J. Woods Halley of the Physics Department. Their work measured helium collisions by placing resistive chromium source mounts on one end of a box and titanium bolometers on the opposite end. It was expected that if the graph of power through the chromium source mounts had a single peak, the graph of power through the bolometer would also have a single peak. However, this was not the case. Investigations of the effects of collisions were made to see if they could account for observed multiple extrema. The simulation was coded in fortran and has been run one thousand times for one hundred particles and ten thousand time iterations. These simulations were run to obtain an average in hopes of suppressing the magnitude of statistical fluctuation as far as possible. From the simulations ran, it was discovered that the working scheme was not as straight-forward as it was hoped.
 |
| Sara Firl prepares for her presentation.
|
 |
| Thomas Miller (left) talks with Jocelyn Rodgers (right) at the program's kickoff reception.
|
Benjamin Miles, a Biochemistry major at Indiana University, worked with Professor William Gleason of the Laboratory Medicine and Pathology Department. Benjamin's work helped develop a three-dimensional model for the Vascular Endothelial Growth Factor (VEGF) 165 isoform. Using the macromolecular modeling programs insight and discover, the known structures of VEGF were joined, and unknown components were modeled. An energy minimization calculation was then performed on the completed structural molecule. Once complete, procheck was used to evaluate it. In addition to the computational work in developing a model of VEGF, some preliminary experiments aimed at developing a quantitative VEGF assay were done.
Sara Firl, a Chemical Engineering major at the University of Minnesota, worked with Professor Alon McCormick of the Chemical Engineering and Materials Science Department. Together, they studied how molecules group or cluster together using various tools such as Monte Carlo simulation. Sara helped with a series of fortran codes used to set up a simple lattice of cubic unit cells with established periodic boundary conditions. The energy of the unit cell was initially calculated using a Lennard-Jones potential. One random molecule was moved, and the energy was calculated after the move occurred. The program ran until a reasonable equilibrium was reached. An energy curve and radial distribution function were plotted from the final structure data. Further development of this project will study the adsorption of xylene molecules into the pores of compounds such as zeolites and clatharates.
Brent Grocholski, a Physics major at the University of Minnesota, worked with Professor George Wilcox of the Pharmacology Department. Brent and Professor Wilcox worked on the development of improved neuronal simulators for sensory nerves. Brent helped write a program that used the Hodgkin-Huxley model and a fourth order Runge-Kutta integrator to simulate neuron action potentials. A program was then written to both approximate the Jacobian for the Hodgkin-Huxley equations and perform an LU factorization used to solve a vector used in DASPK (Differential Algebraic Solver with Preconditioned Krylov methods). The preconditioning matrix failed to reduce the number of iterations needed to solve a time course when the full Jacobian was used, but setting the first row of the Jacobian to zero reduced the number of iterations required to solve the time course by a significant amount.
| Summer 1999 Tutorials |
- 64-bit Parallel MPI and Math/Numerical Libraries for the IBM SP
- Code Optimization Workshop
- Introduction to InsightII/Discover
- Introduction to Molecular Animation
- Introduction to Parallel Programming
- Introduction to Parallel Programming on the IBM SP Supercomputer
- Introduction to Parallel Programming on the SGI Origin 2000 Supercomputer
- Introduction to Perl
- Introduction to the POWER3 Architecture and Code Tuning Guide
- Introduction to Scientific Visualization
- Introduction to Shell Programming
- Introduction to the Supercomputing Institute
- Molecular Visualization Tools
|
|
Charles Park, a Mathematics and Physics major at Yale University, worked with Professor Thomas Jones of the Astronomy Department. Charles helped produce synthetic astronomical observations of magnetohydrodynamical simulations of the jet flow in radio galaxies. These synthetic observations are useful for studying the jets because they allow the study of emissions from relativistic electrons in the jet and the test of reliability of actual observations of radio galaxies and the properties that observational astronomers infer from radio telescope images. The synthetic observations were produced by a raytracing program. By using standard astronomical tools to analyze the synthetic images, results of three-dimensional simulations were made more accessible to observational astronomers. The synthetic images suggested many of the quantities radio astronomers had infered from actual observations. Other values were found to rely noticeably on the orientation of the object with respect to the observer.
 |
| Anna Saputera awaits the start of a tutorial at the Supercomputing Institute.
|
 |
| Amy Beukelman (left) discusses her research with a researcher at the Supercomputing Institute.
|
Thomas Miller, a Chemistry and Mathematics major at Texas A&M University, worked with Professor Donald Truhlar of the Chemistry Department. Together with graduate student Michael Hack, they implemented and tested the Army Ants method, a statistically improved form of the fewest switches trajectory surface hopping method of calculating reaction dynamics for three-atom, two-state systems. A number of analysis programs were written to interpret the new output, and a Monte Carlo simulated trajectory surface hopping program was written to optimize the code's various input parameters. The programs were applied to the collision of an electronically excited bromine atom with a hydrogen molecule, a problematic system with very weak coupling between the two potential surfaces.
Jocelyn Rodgers, a Chemistry and Physics major at Harvard University, worked with Professor Truhlar on optimizing molecular geometries via multi-level linear combinations of electronic structure calculations. Recently, Professor Truhlar's group has developed several single-point energy methods that yield more accurate energies without a prohibitive increase in computer time. This project expanded these linear combination methods to include multilevel optimization of molecular geometries. A code called multilevel, written in fortran90, was developed to calculate the energy, gradient, and Hessian (a multi-dimensional matrix of second derivatives) for a molecule with the multilevel methods. A few molecules have been optimized with the methods in multilevel, and the method shows great promise.
 |
| Joseph Cooley (left) and Brent Grocholski (right) await a tutorial at the Institute.
|
 |
| Professor William Gleason (left) of the Laboratory Medicine and Pathology Department with intern Benjamin Miles (right) before Benjamin's talk.
|
Christine Tratz, a Chemistry major at the University of Oklahoma, worked with Professor Truhlar and graduate student Patton Fast on multi-coefficient correlation methods (MCCM) and multi-coefficient gaussian methods (MCGx) for the efficient and accurate calculation of potential energy surfaces. Investigations were made on the reliability of these methods for types of compounds to which they have not previously been applied. The methods were improved by optimizing them over a larger training set. In order to create a larger test set for the MCCM and MCGx methods, a number of gaussian jobs were run. Spreadsheets that provide the electronic energies were created with the results of the gaussian calculations. The errors in the different methods were then compared. Coefficients for each of the methods were reoptimized and modified. The mean unsigned error in the new methods were shown to be considerably lower than in the previous methods for all cases‹before and after reoptimization.
Aaron Miller worked with Professor Timothy Ebner of the Neurosurgery and Physiology Department. Aaron's work was part of the Computational Neuroscience Internship program. Aaron and Professor Ebner performed experiments that investigated whether the visuomotor pursuit tracking error in monkeys would be a function of target width and speed. A female rhesus monkey was shown to use a two-joint manipulandum to make visually guided arm-tracking movements in the horizontal plane. For each successful trial, x and y position points were recorded, smoothed, and digitally differentiated to obtain speed points. Position error was calculated as the difference between the x and y positions of the hand and target. Speed error was calculated as the difference between the speed of the hand and the target. Position error was not found to change significantly with changes in target size and speed. Speed error was not found to change significantly with target size, but did increase significantly with increasing target speed. Another experiment, in which the target underwent an abrupt two-fold change in speed, found an observable transient increase in speed error.
Karis Stenback worked with Professor Robert Miller. Karis' work was also part of the Computational Neuroscience Internship program. Together, these researchers worked on elucidating the mechanism by which active dendritic spikes (spike initiation) occur, observing how they propagate once they are present, and investigating the functional differences and attributes of active dendritic spiking versus active somatic spiking. Both an equivalent cylinder (EC) model, used to exhibit amacrine cell properties, and a real cell model, created from a trace of an on-off amacrine cell obtained during experiments, were used. Using the parameters from the EC amacrine cell model in the real cell model, alpha synapses were dispersed throughout the dendritic tree in proximal locations. Multiple voltage recording sites were placed throughout the dendritic tree and in the soma to elucidate where dendritic spikes were initiated. To look at the functional differences and attributes of dendritic spiking versus somatic spiking, only the calcium area was looked at. It was found that the intracellular calcium concentration showed a ten-fold increase when the spike mechanism was present and showed almost no change when the spike mechanism was absent.
 |
| Thomas Wilson prepares his work in one of the offices at the Supercomputing Institute.
|
 |
| Sekar Velu (left) and Charles Park (right) before a tutorial.
|
Amy Beukelman, a Chemistry and Biology major at the University of Minnesota, worked with Professor Christopher Cramer of the Chemistry Department. Together, these researchers studied singlet-triplet gaps and transition states. After gas-phase calculations were performed, calculations were carried out in solvents, as the methanol environment chosen more closely resembles experimental environments. Because some of the structures chosen were ions, it was important to look at the structures in solution, where ionic forms of molecules are more favored. Single point energy calculations of the optimized structures were then performed using the amsol program and methanol as a solvent. The gamesol program was used (with methanol as a solvent) to obtain geometries of the molecules in solvent. While the triplet was found to behave classically, the singlet ion rearranged in the gas phase. The singlet-triplet gap was actually quite large indicating that intersystem crossing between the two related molecules was unlikely.
Sekar Velu, a Computer Science and Chemistry major at Syracuse University, worked with Professor Douglas Ohlendorf of the Biochemistry, Molecular Biology, and Biophysics Department. Sekar helped build a computer program to take two undocked proteins and accurately predict their docked structure. The project involved writing a shape complementarity program in C. This program utilized an earlier algorithm in order to generate a correlation coefficent between two protein surfaces. It was then decided that the best approach to take in determining the structure was through the use of the Fourier Transform method. Extensive tests were run on FTDock to determine the most effective parameters. The generated complexes were then tested against the crystallographically determined structure. Electrostatics were found to play a diminutive role in complex conformation, and solvent accessibility was found to be a much more important factor. Shape complementarity was found to play an important role because there appears to be an ideal value for a specific protein-protein complex.
Joseph Cooley, a Computer Science major at the University of Minnesota, worked with Professor Linda Boland of the Physiology Department. Joseph and Professor Boland optimized a model for ion channel gating. A model that simulated ion channel current flow was programmed by allowing the model to input biological data, performing serial and parallel simulations, and normalizing simulation data with biological data. Multiple simulations were then performed via parameter changes, and the resultant simulation with the least error in comparison to biological data was chosen. Finally, multi-platform development was done. Additional work on the model added more optimizations and features.
 |
| Sara Alfano discusses her work.
|
 |
| Computational Neuroscience interns Aaron Miller (left) and Karis Stenback (right) await the start of presentations.
|
David Dreytser, a Chemical Engineering and Chemistry major at the University of Minnesota, worked with Professor David Thomas of the Biochemistry, Molecular Biology, and Biophysics Department. They worked on computational simulations of electron paramagnetic resonance (EPR) on Phospholamban, a 52 amino acid integral membrane protein of the sarcoplasmic reticulum. The three-dimensional structure of phospholamban was extended into the cytoplasmic domain. Initially, a script was written to convert formats, and a spin label was built. This spin label was attached to various binding sights on phospholamban before molecular dynamics were performed.
Anna Saputera, a Computer Science major at the University of Minnesota, worked with Professor M. Germana Paterlini of the College of Pharmacy. Anna and Professor Paterlini calculated the electrostatic free energy of interaction of an opioid peptide as a fuction of distance and orientation from a lipid bilayer. Anna helped create a C++ program to rotate and transform the position of the peptide molecule with respect to the lipid. She then used midas for visualization of the system before and after the coordinate transformation. The electrostatic potential energy of the molecule/lipid systems were calculated at different translation and orientations and with different input parameters for delphi to check the performance of this program. The potential energy between molecule and lipid was then calculated as a function of distance. The profile showed an atractive interaction energy between the lipid and the peptide. The energy profile obtained using delphi was checked for convergence using different input parameters.
David Liebelt from Northwestern University worked with Professor David Rottenberg of the Neurology and Radiology Department at the VA Medical Center. This work was part of the Computational Neuroscience Internship Program. David helped create a World Wide Web interface for the Corner Cube Environment, which faces several hurdles before becoming common neuroimaging software. The interface consists of html documents displayed in an html frameset. With JavaScript, this interface is able to store image data and viewing preferences chosen by the user. With the data and preferences stored, JavaScript generates an html form that the user can submit over the Internet. The form is processed by a cgi script written in perl. This processing involves interpreting the data and preferences, writing data and preference files readable by the Corner Cube Environment, running the Corner Cube Environment, and displaying the output to the World Wide Web browser.
Cody Zilverberg, a Computer Science major at St. John's University, worked with Professor David Lilja of the Electrical and Computer Engineering Department on performance visualization tools for java programs. Coding was done on the 'Find Next' component of a program called JaViz. JaViz visually displays a tree of nodes where each node represents a method call in a Java program. 'Find Next' contains a simple graphical user interface (GUI) that allows users to find the next node in the JaViz tree that matches certain criteria. Because there can be many thousands of nodes in a single tree, an algorithm needed to be developed that would read node data from a file, examine the data, and then (for memory purposes) discard unnecessary data. This search was implemented by using a background thread that allowed the visualizer to continue functioning during execution.
Thomas Wilson, a Computer Science and Architecture major at the University of Minnesota, worked with Professor David Yuen of the Geology and Geophysics Department on a two part project. The first part of the project consisted of a World Wide Web presentation focused on scientific visualizations related to nuclear waste transportation and processing at the Hanford, Washington nuclear waste reserve. This report consists of critical reviews and improvements made to the visualizations. The second part of the project involves Java. Tom helped implement a generic Java applet control panel that included useful features for the Java based pV3 graphical user interface (GUI). The features included mathematical operations initiated by a mouse click, writing to the standard output, and providing an echo area in the window. Overall, this work found Java to be a good language to use for the new version of pV3 because of security, functionality, portability, and the GUI widgets needed.
Christine Tratz
(left) and Maegan Harris (right), an undergraduate researcher in Professor
Truhlar's Chemistry group.
|
Cody Zilverberg (left) and David Dreytser
(right). |
| Summer 2000 Program |
The Supercomputing Institute is pleased to announce its Undergraduate Internship Program for Summer 2000. Summer appointments will be full-time, ten-week appointments. The 2000 program will run from June 12 through August 18, 2000. A student interested in becoming an intern must be an undergraduate student at the time of the internship to be eligible and must be a citizen or permanent resident of the United States and its possessions.
All applications are judged competitively based on the qualifications of the applicant and the availability of a suitable project.
Application forms, appointment information, and project lists are available on the World Wide Web at:
www.msi.umn.edu/general/Programs/uip/uip.htm
Application forms and project lists are also available by contacting:
Undergraduate Internship Coordinator
University of Minnesota
Supercomputing Institute
1200 Washington Avenue South
Minneapolis, Minnesota 55415-1227
Phone: (612) 626-7620
Email: uip@msi.umn.edu
|
|
|
|
This information is available in alternative formats upon request by
individuals with disabilities. Please send email to
alt-format@msi.umn.edu
or call 612-624-0528.
|
|
URL: http://
|
|
|
This page last modified on
|
|
Website related questions or problems should be directed to
webmaster@msi.umn.edu
The Supercomputing Institute does not collect personal information on
visitors to our website. For the University of Minnesota policy, see
www.privacy.umn.edu.
|
|
| |
