University of Maryland

The Right Tools for Research

Landavere says his multiplatform approach to computing infrastructure derives from his experience watching IT people trying to hammer technical square pegs into round holes.

“I think some colleges limit their effectiveness by limiting the platforms they support,” Landavere says. “It’s like you come into a group and drop off a saw, a hammer and a bag of nails. As you walk out, you ask ‘By the way, what do you guys do?’ And they say ‘We grow crops.’

“We will never do that. When we meet a research group or work with a new faculty member, we start with questions like ‘What is your purpose? What does success look like?’ Then we align platform, applications and IT solutions with their mission.

“When you find the right tool for the right job, the support effort is easier,” Landavere says. “It’s smoother. It’s more seamless.”

Mac OS X Powers the Entire Workflow

“Because Mac OS X is based on UNIX, the entire open-source world is available to us and to our research groups,” Landavere says. “Porting open-source applications to the Mac has been fairly trivial, and all of a sudden, on one platform, our researchers can run their open-source scientific applications and have productivity, email, Word documents.

“All of a sudden, on one platform, our researchers can run their open-source scientific applications and have productivity, email, Word documents.”

“For most of the folks who use Macs, it’s the only platform they need. They do everything — research, data gathering, analysis, publications and communications — all on one machine. That’s really nice.”

An Open Platform and Philosophy Spur Innovation

Today, the IT group in the College of Life Sciences meets regularly to collaborate on technical issues and successes within their departments, across the college and throughout the University of Maryland.

“In a diverse, multiplatform environment, we have to work as a team to be effective,” Landavere says. “One of our IT group members might say, ‘Hey, one of our faculty members needs to run an image analysis app, and he’s not sure which one to use.’

“Someone else will say, ‘A new faculty member in my department just set up shop with a couple of Power Mac G5 systems and uses XYZ application for image analysis. Let’s put them together.’

“That’s where true innovation and excitement happens,” says Landavere. “The open-platform and open-source philosophy of the IT group and the open-source philosophy of the Mac platform are a perfect fit.”

1 2

Macs in Life Science Research

Cell Biology and Molecular Genetics

Cell biologist Dr. Stephen M. Wolniak is a professor and associate chair of the college’s Cell Biology and Molecular Genetics department. He combines RNAi strategies and bioinformatics tools on Macs and an Xserve G5 cluster with high-end microscopy to study the process of cell fate determination in plants and the mechanisms that control patterns of gamete development.

“I moved to the Xserve because we can port the images we capture from the microscopes directly to the server,” Wolniak says. “The cluster nodes run algorithms that clean up the images and perform some online filtering. We store both the raw and processed files on the RAID, even as subsequent images are acquired during the course of experiments.”

Reconstructing Plant Evolution

Dr. Charles Francis Delwiche, associate professor of Cell Biology and Molecular Genetics, uses genomics techniques and bioinformatics to reconstruct the history of plant evolution.

“The Macs are terrific for all the kinds of work we do,” he says. “Sequence analysis requires high performance, and we have a lot of graphical work that involves acquiring, tracking and processing images from the microscope, preparing them for publication, and publishing them on the web.”

Porting Molecular Dynamics Codes to the Mac

Along with his team of researchers, Dr. Millard H. Alexander, distinguished professor in the Department of Chemistry and Biochemistry, focuses on simulating the outcomes of basic chemical reactions. Alexander and his team pursue their work running large-scale numerical calculations on Power Mac G5 systems and an Xserve cluster.

“We particularly enjoy the G5 processor, which is very fast for a number of these applications,” Alexander says. “We write our codes in Fortran, and Macs use the same sophisticated Fortran compiler as the IBM RS6000 platform, on which many of these codes were originally developed. So it’s very easy to port our codes to the Mac.”