The cornerstone of the Ohio Supercomputer Center's Summer Institute is the projects. The students work together in small teams on diverse and challenging research-level projects. Teams are comprised of a project leader (staff member who conceived and designed the project) and three or four students.
This year's project options are:
1) Lab-on-a-chip Nanofluidics - more info
The goal of new lab-on-a-chip nanotechnology is to shrink today’s diagnostic equipment onto chips that are about a million times smaller. In the medical field, this offers the possibility of diagnostics based on ultra-small samples, single-cell analysis, and portable devices for rapid and accurate bedside diagnosis of conditions like heart disease and cancer. An ordinary pathology lab involves mixing samples and chemicals, and moving them into position for analysis. These same functions must be performs on lab-on-a-chip devices. Therefore, we need reliable methods for transport of fluids and biomolecules in nanoscale structures.
You will explore one of the most promise methods of fluid transport in nanostructures, electroosmotic flow (EOF). The principle behind EOF is simple: Most surfaces carry an electrical charge. That means there must be charge of the opposite sign somewhere to compensate the surface charge. For example, ordinary glass has a strong negative charge, and there is always a compensating layer of positively-charged fluid containing ions like Na+ or K+ near the surface. Application of a voltage to the positively-charged fluid causes fluid flow known as electroosmotic flow (EOF). The advantage of EOF is that fluid flow is under precise electrical control without the use of mechanical pumps.
You will explore how surface charge, surface roughness, and salt concentration affect electroosmotic flow, and draw conclusions about the conditions for optimal efficiency. You will compare detailed molecule simulations with continuum models of fluid flow like the Navier-Stokes equation. (Click on the image above to see an example animation)
2) Image Processing - More Info
This project involves a real world application of finding comets in sun observation images from the SOHO (Solar and Heliospheric Observatory) spacecraft.
SOHO is a cooperative mission between the European Space Agency (ESA) and the National Aeronautics and Space Administration (NASA). SOHO studies the sun from deep down in its core out to 32 solar radii. The spacecraft orbits the L1 Lagrangian point. From this orbit, SOHO is able to observe the sun 24 hours a day. Even though SOHO's primary objectives relate to solar and heliospheric physics, the onboard LASCO instrument has become the most prolific comet discoverer in history!
LASCO (Large Angle Spectrometric Coronagraph) is able to take images of the solar corona by blocking the light coming directly from the Sun with an occulter disk, creating an artificial eclipse within the instrument itself. LASCO images are automatically posted on the web approximately every 20 minute. Since LASCO began taking observations in January of 1996, the C2 and C3 coronagraphs have observed over 950 new comets and 9 known comets. The vast majority of these comets were discovered by amateur astronomers who closely examine the images for potential comets. Below is a typical image recently taken by SOHO.
3) Cancer Cell Migration, Invasion and Metastasis
Cancer is a major cause of death in the United States and the spread of cancer cells from a primary tumor to other organs, i.e. metastasis, is the key feature that leads to high mortality. For example, women with localized breast tumors have a 98% 5-year survival rate. However, metastasis of local breast tumors to other organs leads to a very low 23% survival rate. Metastasis involves the detachment of epithelial cells from the primary tumor, cell migration into the surrounding tissue, cell invasion into blood/lymphatic vessels and colonization of distal organs. One biological mechanism responsible for metastasis is known as epithelial to mesenchymal transition (EMT). EMT is a form of cell-plasticity (i.e. stem cell like behavior) where non-motile epithelial cells get converted into a highly migratory/invasive mesenchymal cell. EMT is normal assessed by measuring the activity of certain genes that are modulated during the transition. Unfortunately, many of these gene markers are non-specific and do not undergo consistent changes during metastasis. Furthermore, these markers do not provide a quantitative way to determine the degree of EMT and/or the metastatic potential in a given patient.
Recently, our laboratory has demonstrated that cancer cells undergo dramatic biomechanical and structural changes during EMT (see Figure 1). Not only are these biomechanical changes are highly-specific, since characterizing cell/tissue mechanics is a clinical viable diagnostic technique, these changes in cell mechanics represents a novel way to quantitatively assess the degree of EMT. However, it is not well understood how changes in cell mechanics facilitate or alter the cell migration and invasion processes required for metastasis. We have therefore started to develop sophisticated multi-scale computational models to investigate how changes in different cell biomechanical properties influence cancer cell migration and metastasis (see Figure 2). These models can simulate both the detachment of cancer cells from the primary tumor and their migration/invasion into surrounding tissues. For the summer projects, students will first be exposed to the computational tools used to create models of cancer cell migration and invasion (i.e. finite element modeling). The student team will then be ask to model a very specific and important step in metastasis where cancer cells that have detached from the primary tumor must squeeze through pores in the extracellular matrix and invade through a thin layer of endothelial cells to enter the blood stream. Two PhD students from Dr. Ghadiali’s lab will be available to help students set-up initial models and analyze model results. These students will also describe how they are using computational techniques to advance their PhD research programs.
Sponsored in part by NSF grant 1134201
4) High Performance Programming
Supercomputers from the 1990's to today are built as cluster systems. A cluster is a collection of individual computers called "nodes" (each like a PC) that are collectively programmed to solve a single problem. From 1990 to 2004, supercomputer power increased due to two factors: first, increased processor speed (the nodes got faster) and second, increased cluster size (more nodes in a cluster). Now, both of those factors have ceased: processor speeds have been flat since 2004 and cluster sizes are constrained by cost and power consumption. Instead of larger and faster, supercomputers are now getting denser, with additional processing cores and accelerators like GPUs within each node. These changes are only beginning and will cause major changes to the way we program our largest machines.
Effective programming for mulitcore nodes requires good performance on a single core along with efficient distribution of work across cores. Students will learn basic techniques for improving single-core performance, the OpenMP library for work distribution and offload libraries for accelerators. Students will then design a project focused either on performance measurement and improvement for an existing code or creation of a parallel program to use multiple cores or accelerators. It is required that students who chose this project have already have some experience coding in C or C++.