Lafayette Campus News (www.lafayette.edu), June 15, 2005 — Rob Root, associate professor of mathematics, and Chun Wai Liew, assistant professor of computer science, are half of a team that has been awarded a $956,000 grant from the National Science Foundation to study how vertebrae first evolved. In the process they are training groups of undergraduates how to constructively engage in scientific research that is cross-disciplinary.

The grant provides Lafayette students with opportunities to build mathematical models and computing frameworks that serve as useful tools for biologists investigating the origin of vertebrae and other questions surrounding fish locomotion. The complexity and multidisciplinary nature of the project will enlist student researchers throughout the four-year duration of the grant.

The project teams Root and Liew with biologist John Long of Vassar College in Poughkeepsie, N.Y., and biomedical engineer Tom Koob of the University of South Florida and Shriner’s Hospital for Children in Tampa, Fla.

“The four of us are creating models, both robotic and digital, that are biominetic (imitating biological systems),” reports Root, who has studied many aspects of fish locomotion since 1996.

“The question we’re trying to answer is ‘why did vertebrae evolve in fish?’ We know ancient fish had a spinal cord and a surrounding structure, a notochord, before they had backbone. So why bother with a backbone? What advantages do vertebrae offer?” asks Root.

“This transition occurred in incremental fashion; the vertebral column didn’t change abruptly from no bones to the many bones with intricate geometry and complicated joints that we observe in present-day fish. There were steps along the way and each has to have a reason,” adds Liew.

“The backbone is a critical feature in the eventual evolution of amphibians and terrestrial vertebrates, and in humans’ development of the ability to stand erect. But where do [vertebrae] come from? The biologists we work with tell us that’s a big question,” says Root.

In creating virtual fish and modeling them, the team hopes to offer plausible answers to why vertebrae evolved based on swimming performance. Few fossil records of primitive fish exist, although recent excavations in China have uncovered records of transitional species that might lead to further insights.

But the paucity of fossil remains means the structure of these fish present the project team with “a jig-saw puzzle with incomplete pieces,” says Liew. “We’re trying to figure it out, but you have to make some guesses along the way, and it’s easy to go wrong.”

Root and Liew are working together to create digital models, or simulations, for the project. “We’re trying to create more realistic physiology because the difference between a notocord and backbone is fairly refined,” says Root. “To answer biologically significant questions we must know exactly how those muscles right next to it worked. So we need better models for the neurostimulation of the muscles, more realistic models for the fluid dynamics. The result is a model that will behave very differently based on our assumptions about the fish and its environment. For example, exactly how strong are the muscles? How flexible are the joints? How stiff are the bones? Each of these is a number, a parameter, for the model. Some parameters our biologist colleagues can measure experimentally, but others we can only guess at.”

Root models fluid dynamics and analytical mechanics while Liew seeks to optimize the myriad parameters through efficient computation.

“We start with a simplified model of a fish that has approximately 20 variables,” says Liew. “In the full form, we will eventually have 80 to 100 variables, perhaps more. The values of the variables affect the swimming behaviors for the fish: swimming faster, turning more effectively, etc. We have digital video of actual fish, and what we want to do is take the model and find the values of the parameters that will get us the swimming motion that closely resembles the real fish. This is an optimization problem with a very large number of parameters and therefore requires both efficient algorithms and a great deal of computing resources. ”

The basis for the modeling will be work Root has done over the past nine years, research that often has been done with students. His collaborations have involved a variety of majors, such as Myat Lin '04 (Yangon, Myanmar), biochemistry; Pujitha Weerakoon '04 (Kandy, Sri Lanka), electrical and computer engineering; and Bruce Adcock ’01 (Watervliet, N.Y.), mathematics. Root has also incorporated undergraduates from other institutions in his research through the mathematics department’s Research Experience for Undergraduates, which is also funded by the National Science Foundation, and draws students interested in mathematics research from all over the country to Lafayette every summer.

The optimization work is based on the framework that Liew has developed in collaboration with EXCEL Scholars Mayank Lahiri ’05 (Muscat, Oman), Daniel Huber ’02 (Philadelphia, Pa.), and Matthew Patton ’02 (Los Alamos, N.M.).

Lafayette’s EXCEL Scholars program, which has helped make the College a national leader in undergraduate research, will give an average of three students per year the opportunity to work with Root and Liew. EXCEL students assist faculty with research while earning a stipend. Many of the more than 160 student participants each year go on to publish papers in scholarly journals and/or present their research at conferences.

“The multidisciplinary aspect [of the project] makes it so beautiful for undergraduate work,” says Root. “There are so many pieces of this that you don’t need to be a graduate student to participate — an undergraduate just needs to be able to contribute to one aspect of the larger model; that’s how we all start out in research.”

Another benefit to the Lafayette community is a substantial increase in scientific computing resources. The grant allocated $27,000 for computers, which was augmented with $55,000 from Lafayette. Sixteen high-speed computers purchased for the project will be used for other science and engineering projects requiring high-performance parallel computing.

This is the second National Science Foundation grant in two years received by a multidisciplinary team that included Root. He and several other Lafayette professors received a $243,526 grant establishing a engineering laboratory providing new learning opportunities in several areas for mechanical engineering, civil and environmental engineering, and mathematics students.

Other recent grants received by Lafayette professors supporting their research with students include: