New Tool for Genomics Research and Education Works Like MAGIC
Laurie Heyer and Malcolm Campbell and student assistants created MAGIC Tools for genomics education and research.
|
8/11/2003
Contact: Bill Giduz 704/894-2244 or bigiduz@davidson.edu
Genomic science promises to revolutionize medicine by individualizing medical treatment to match a patient’s DNA. Malcolm Campbell, a genomics specialist at Davidson College, predicts that labs will measure the activity of genes in a patient’s healthy cells and compare it to the activity of genes from a diseased cell. The results will show doctors which genes are “out of kilter,” leading them to prescribe medications that specifically address a patient’s condition.
Campbell, an associate professor of biology, published the first-ever genomics textbook, Discovering Genomics, Proteomics, and Bioinformatics, last year with his Davidson colleague, Laurie Heyer, an assistant professor of mathematics who specializes in bioinformatics.
A lot of what’s needed is already in place. Individual genes can be printed out on a “microarray.” In essence, it’s a glass microscope slide on which a cell’s genes are displayed as tiny dots and arranged in big rectangles called “arrays.” The problem is that the human body has about 30,000 genes! Analyzing 30,000 dots on a microarray can’t be done with the human eye. That job requires some serious computer power. Some software is available as a commercial product, but costs a mint of money.
Davidson’s genomics team, Heyer and Campbell, recognized that the price tag was a stumbling block for many academic researchers who want to learn and teach the quickly evolving discipline of genomics. However, they have turned the stumbling block into an opportunity to position Davidson as a genomics leader through development of MAGIC Tool, a public domain software package that in many ways analyzes microarrays far better than the expensive commercial software.
Emily Oldham ' 03 devised this interpretation of relationships between genes for MAGIC Tools.
|
“Genomics is all about volume and scale,” said Campbell. “The amount of data that comes in is like trying to drink from an open fire hydrant. You’re looking for what may be just a few key genes among thousands. MAGIC Tool gives you the means to find out which data are significant.”
MAGIC Tool highlights the importance of Heyer’s field —bioinformatics— in the genomic revolution. It employs mathematical algorithms to analyze the voluminous data on a microarray and display it in many different forms to help users draw conclusions. Heyer and five Davidson students spent three summers developing the software with support from grants by the National Science Foundation and the college. None of the students had any previous experience with microarrays, but Heyer said their “refreshing naievete” led them to develop unique, helpful new ways to display and express results.
The Davidson software team faced substantial challenges. One of the first important steps was to make it easy for users to systematically draw an imaginary box around each gene’s dot on the microarray, placing each gene in an identically sized square. Students Danielle Hyun-jin Choi ’04 and Parul Karnik ’04 met that challenge. Then it was necessary to locate the gene’s dot in the square, and quantify the “brightness” of each. Genes appear as brighter or darker according to whether they are active in the biological process under study (e.g. cancerous vs. non-cancerous tissue). But the spots appear as many shades of gray, and occupy varying amounts of space inside their gridded squares. All those variables can affect the interpretation of data, which means that the software must be both accurate and adaptable.
Once measurements are made, researchers can employ a variety of methods to analyze the data. Adam Abele ‘03, who worked on the project for two years, developed a means to display the data as a dynamic table of ratios. His displays allow the user to easily detect each gene’s activity by adjusting the intensity of the table’s colors to more easily recognize patterns. Heyer commented, “Adam didn’t know that no other software had this type of display, he just came up with it by himself. It’s a great tool, and unlike any other one out there.”
A database associated with the program shows users the name of individual genes, the chromosome on which they appear, and their biological and molecular function, by clicking on the gene. Users can search and display data that meet certain criteria, and plot those results in many different ways.
Each square on the microarray represents a gene, and its intensity reflects the gene's activity.
|
Genes that are behaving in similar ways can be clustered together to facilitate pattern recognition. A “clustering component” displays a hierarchical “tree” of the genes. Users can click on one gene and highlight the lines representing the first gene’s relationship to other genes. Another innovative display allows users to highlight parts of tree and “explode” areas for more detail, or “collapse” it for less detail. Emily Oldham ’03 suggested another novel display which places genes in a circle on the screen, with lines connecting ones that meet designated thresholds.
MAGIC Tool solves many of the technological challenges of data interpretation. But Heyer and Campbell said the most important part of the process is the user’s ability to draw conclusions. Campbell said, “MAGIC Tool crunches the numbers, and allows you to visualize the outcome in ways that may give you insights, but the ability to recognize the biologically significant patterns is dependent on an individual’s training and experience.”
To illustrate the importance of human interpretation of patterns, Campbell cited discovery of the cause of cholera. He said, “Queen Victoria’s physician decided to plot out on a map of London where each cholera case occurred. He looked at all his pins and noticed that they were clustered around wells. That led him to conclude, correctly, that cholera is spread by water.
“The ability to look at massive amounts of data and draw correct inferences is what’s important in genomics. MAGIC Tool allows investigators to start working on that in earnest,” he said.
To help encourage the widespread use of MAGIC Tool, the software was created with an “open architecture” that permits users access to the coding of the algorithms, and to modify the code to create their own displays.
That type of open architecture will also help establish its credibility as a reliable standard in the industry. “DNA microarrays are like the Wild West,” said Campbell. “The field is only five years old and people are making up rules as they go along. Everyone’s out to cure cancer, and that’s a good thing. But they’re skipping a lot of important steps in testing and validating the methodology that can lead to misinterpretation of the data.”
Heyer and Campbell are leading four workshops this summer to demonstrate MAGIC Tool. With the help of one of the lead code writers, David Moskowitz ’03, they already demonstrated MAGIC Tool to scientists from the California State University system, where MAGIC Tool will be adopted on several campuses. The fifty members of the Genome Consortium for Active Teaching (GCAT; www.bio.davidson.edu/GCAT), an organization that Campbell helped found to promote undergraduate instruction in genomics, will adopt MAGIC Tool on their campuses across the country, in Canada and Australia. The software is posted on the biology department’s web server for free downloading.
Campbell will score another “first” in genomics instruction as co-recipient of a $1.2-million, three-year grant from the Human Genome Research Institute and the National Institute of General Medical Sciences to create multimedia tools for learning genomics. “Multimedia is a necessary component for genomics instruction,” he said. “There’s still a need for printed material, and for students to ask questions of a live professor, but multimedia can tie together complex processes that are difficult to convey on paper.”
The multimedia tools, which will be created over a three year period, will be posted on a Web site, made available on DVD, and will be bundled with the next edition of Campbell and Heyer’s book, Discovering Genomics, Proteomics, and BioInformatics. Campbell is in charge of the content for the multimedia project, and plans to present all the content as a series of real-life cases that illuminate genomic principles. For instance, he plans to cite mad cow disease as an example of proteomics. Mad cow disease is a “prion” disease caused when a particular type of protein contained in every cow brain changes shape. The case will not only describe the disease, but will show how it spreads through feeding ground-up cow parts to other cows, and through the global nature of the livestock trade.
Campbell noted that genomics puts students almost on a par with their professors, since the tools and discoveries are being released almost every day. “I’d love to be a bio major coming up now,” he mused. “Students can learn a scale and complexity of data that earlier tools couldn’t handle. It’s like being a student at the time van Leeuwenhoek invented the microscope. Students and researchers are getting a first crack at the cutting edge of the science the same time. It’s an incredibly exciting time to be learning—and teaching—genomics.”
Davidson is a highly selective independent liberal arts college for 1,600 students. Since its establishment in 1837, the college has graduated 23 Rhodes Scholars and is consistently ranked in the top ten liberal arts colleges in the country by U.S. News and World Report magazine. Davidson is engaged in “Let Learning Be Cherished,” a $250 million campaign in support of student financial assistance, academic resources, and community life.
# # #
|
