Digital laboratory

A sophisticated computer lab at LSU-Shreveport is helping to turn the flood of data from the life sciences into meaningful knowledge.

In the late-night stillness of a science lab filled with computers and monitors rather than Bunsen burners and beakers, a research scientist develops an algorithm that could-when applied to volumes of data produced by other researchers-reduce the detection time for a virus such as SARS from four days to a matter of hours.

Such is the work taking place today-maybe even tonight-at the Laboratory for Advanced Bioinformatics (LABi) on the Shreveport campus of LSUS.

LSUS provides an academic and research environment where experts from a wide range of fields can talk to each other on a meaningful and productive level. At LABi, for example, the computer scientists learn the ins and outs of biology and chemistry, while the life scientists learn the intricacies of computer science. The results of this cross-pollination of expertise are sophisticated and biologically relevant tools and techniques in an important emerging field called "bioinformatics."

The National Institutes of Health's Biomedical Information Science and Technology Initiative Consortium defines "bioinformatics" as "research, development or application of computational tools and approaches for expanding the use of biological, medical, behavioral or health data."

Bioinformatics experts like Dr. Marjan Trutschl and Dr. Urska Cvek at LABi are using powerful computers to help manage and analyze biological, chemical and medical processes in the interest of improving medicine, from diagnostic techniques to therapeutics to patient information. Their stock in trade is databases: gene and protein sequence data, structure/function information, patient databases, and a wide assortment of data collections from biological research.

The ultimate goal of bioinformatics is to turn these complex and diverse data sets into useful and valuable knowledge that enhances the understanding, prevention and treatment of a variety of diseases-from cancer to diabetes to Alzheimer's and Parkinson's.

Flood of information

The modern life sciences have created a flood of information, dramatically increasing the need for powerful computing tools to help compile, sort and analyze these massive amounts of data. Explorations of the human genome, for example, are generating an unprecedented volume of complex data. By 2003, the Human Genome Project had sequenced more than 400,000 DNA fragments and was believed to be complete. But the next phase involves even more daunting data management problems as scientists begin to compare entire genomes, gene families from different species, and even variations between individuals.

"This wealth of data and information that needs to be sorted through to find the nuggets of knowledge can overwhelm the life scientists," says Trutschl. "At the same time it makes computational scientists like us excited-we get eager to crunch some numbers!"

So, while dramatic changes in medicine might be the ultimate goal of one of their projects, Cvek and Trutschl are working in the trenches: their mission is the development and use of "algorithms," the step-by-step recipes computers follow for program execution.

Through computational analysis with sophisticated algorithms, Cvek and Trutschl can facilitate, promote and harness a variety of interdisciplinary research, making massive databases more meaningful and more useful. At LABi, computer scientists, mathematicians, statisticians, biologists, chemists, medical researchers and students-both undergraduate and graduate-work shoulder-to-shoulder on a variety of research projects, solving real-life problems.

Current research at LABi revolves around collaborative projects between LSUS, the LSU Health Sciences Center at Shreveport, the Biomedical Research Foundation of Northwest Louisiana and the University of Massachusetts. The research ranges from NIH-funded microarray gene expression analysis for protein regulation to computational support for cancer research to the development of novel neural-network-based tools and algorithms for large-scale sequence analysis.

Pictures worth a thousand words

It only takes a simple worm to illustrate how complex the data of the life sciences can be-and how bioinformatics can help. Cvek and Trutschl are currently helping Dr. Robert Rhoads at LSUHSC-S with his study of the soil nematode Caenorhabditis elegans (a small worm about 1 mm long and hardly visible to the human eye), using machine learning and visualization techniques to discover relationships between translation and developmental stage expression of C. elegans based on Rhoads' microarray gene expression studies.

A microarray of the C. elegans worm contains a powerful model of a biological system that has been used for a variety of gene expression profiling studies. (A microarray is a small grid of samples of genes-usually tens of thousands of them-on a surface of around one square inch, which provides data on the expression of the genes.) Rhoads has extended the study of C. elegans to investigate its genes at the level of "translation," the process by which the genetic code is actually used to synthesize proteins, which in turn produce certain types of cells and characteristics.

Cvek and Trutschl used a series of algorithms combined with visualization techniques to identify new, "non-trivial" and interesting relationships that could be extracted from the combined information of multiple microarray expression data sets.

"Visualization is a technique that allows us to visually represent and explore large amounts of data," says Cvek. "Remember, a picture is worth a thousand words, after all!"

"It is beyond the human ability to analyze over 18,000 genes of the C. elegans worm using tables of data-you cannot look at a table and make sense of it," explains Trutschl. "It is a picture that has all this data in it-every gene represented by a dot, for example-that helps us identify patterns and groups of genes. And when combined with statistical information and interactive exploration tools, we get some of the most powerful explorations of the data possible today."

High-speed computing

It doesn't hurt that LABi is the home of "Crayfish," one of the most powerful processing systems in the region. Crayfish is a supercomputer whose computational power places it on the list of the world's Top 500 fastest supercomputing facilities. The cluster was built with a combination of LSUS and external grant funds. The multidisciplinary faculty team involved in the project has been successful in securing external grant funds for equipment, student research fellowships and research project support from a number of sources, including the National Science Foundation, the Louisiana Biomedical Infrastructure Network, the Biomedical Research Foundation of Northwest Louisiana, the Environmental Protection Agency, the U.S. Department of Agriculture, the Louisiana Board of Regents and others.

The potential impact of Crayfish is enormous, since it can be used for a variety of applications-not only bioinformatics and cheminformatics, but also modeling, imaging, prototype design, geospatial analysis and interdisciplinary science and engineering.

Crayfish is also making a big impact on learning at LSUS, complementing many existing resources across a range of academic programs. It not only enhances the coursework of science, engineering and technology majors, but it also promotes development of interdisciplinary courses. "Both students and faculty stand to benefit from access and exposure to the advanced computational analysis environment Crayfish brings to campus. It provides the real-world, high-performance infrastructure that will elevate LSUS's educational and research capabilities," predicts Trutschl.

Image immersion

For LABi, the value is obvious. A high-speed computer like Crayfish is critical to making the most of LABi's other software and computers, including the stereoscopic projection system that serves as the facility's primary visualization device.

In a stereo projection system, two images are rendered, one for the right eye and one for the left eye. These are projected from two projectors onto the same area on a screen, so that they almost overlap. The light from each projector is polarized using linear polarizing filters mounted on each projector (identical to a polarizing filter on a camera) and projected onto a special polarization-preserving projection screen. The stereo image can then be viewed from anywhere in the room by wearing a pair of theme park-style polarized glasses that allow the left eye and right eye images to be processed separately. The two images are then combined in the brain, creating a three-dimensional image.

"This enables us to display three-dimensional objects with a depth of vision that is not available with other projection displays or computer screens," notes Cvek. "We can actually interact with objects such as molecules in an immersive environment."

LABi also serves as a backbone for courses in bioinformatics and cheminformatics taught at LSUS and LSUHSC-S, helping to equip students with a competitive advantage in the job market and businesses with a highly skilled work force ready to tackle the most challenging tasks. [Currently, Cvek and Trutschl work with a dozen undergraduate and graduate students on all aspects of their research, from the hands-on maintenance of the Crayfish cluster and the calibration of the 3-D projection system to writing software for 3-D projection, image analysis and genetic sequence investigation.]

Visualizing the future

It is no coincidence that a facility like LABi has developed at LSUS, as Shreveport has always been a stronghold of medical and health-related research and is a regional health care center for northwest Louisiana and the Ark-La-Tex. There are 21 hospitals and more than 1,200 physicians within the three-parish metro area.

Shreveport's biotechnology cluster is both an economic engine for business growth and key building block in Louisiana's long-term economic development strategy, Vision 2020. Working closely with LSUHSC-S, a wealth of top-quality medical facilities and world-renowned research scientists, LABi is at the heart of this cluster, and sophisticated information technology like that being advanced by Cvek and Trutschl plays a critical role in the development of other Vision 2020 tech sectors as well.

As new sources of biological data continue to emerge, multi-scale and large-scale applications from the LSUS bioinformatics research and development program will play a critical role-perhaps, one day, even leading to better treatments for cancer, Alzheimer's, Parkinson's, and SARS.

Says Cvek, "We're computer scientists at heart, but knowing that our applied research could help find the cure for cancer or give insights into some other deadly disease gives us tremendous satisfaction. That's the appeal of bioinformatics."

Compatible interfaces

How likely is it that two scholars, both natives of the same tiny central European country with doctorates in computer science, would work together at a university in Louisiana to build a research facility and focus on bioinformatics? It wouldn't take a supercomputer to calculate that the probability is about as remote as winning the lottery-at least until you learn their stories.

The homes of Drs. Urska Cvek and Marjan Trutschl are about two hours apart in their native Slovenia, a country of 2 million people which lies at the heart of Europe where the Alps meet the Adriatic Sea. It wasn't until they were both college students in the U.S., however, that they met. And that meeting, itself, was pretty improbable.

Trutschl originally came to the U.S. as a high school exchange student. After returning to Slovenia he worked for four years as a tour operator and ski instructor while he studied mechanical engineering at a Slovenian university. But his real love was computers, having gotten his first one as a teenager in 1981.

In 1991 he returned to the U.S. to study computer science. He stayed with his friends, attended LSUS and graduated in 1994 with a B.S. in computer science.

Cvek came to the U.S. in 1993 as a senior-year exchange student at Indiana University. She studied for a year in Bloomington, then returned to Slovenia to complete the thesis for her bachelor's degree of economics at the University of Ljubljana in Slovenia's capital. She received her degree in 1995 and returned to the U.S. to pursue an MBA at the University of Massachusetts Lowell after turning down full tuition with expenses paid at Hofstra University. At that time Trutschl was already a doctoral student in computer science in Lowell.

They met when Cvek was in Bloomington and Trutschl was in Shreveport. There wasn't much on the Internet at that time, but both had access to it. Noticing Urska's introduction when she signed up for a Slovenian mailing list, Marjan saw that he had worked close to her hometown and contacted her. They got to know each other through e-mails, telephone conversations and several visits from Shreveport to Bloomington-almost 10 times the driving distance between their Slovenian hometowns.

Cvek finished her MBA in 1997, switched her focus and two years later switched to the doctoral program after fulfilling the requirements for the master's degree in computer science at UMass. It was at the encouragement of a professor "you couldn't say no to" that she sought and earned an Sc.D. in computer science in 2004.

Both completed doctoral dissertations in the fields of visualization, data mining and large-scale data analysis-simply stated, the backbone of bioinformatics. Trutschl joined the LSUS faculty in fall 2002 and with Cvek's involvement established the LSUS Laboratory for Advanced Bioinformatics. Cvek joined the faculty after completing her doctorate in 2004.