Computer science graduate develops models for DNA sequence analysis

By Marc Speir
University News Service
April 25, 2007

Leaning over the podium at the College of Science undergraduate commencement at Texas State University-San Marcos in August of 2005, Evan Huston spoke about the importance of education.

Selected as speaker in part due to the work on his thesis, Biometric Security: Speaker Verification with Hidden Markov Models, the computer science graduate and Austin native encouraged graduates to contribute to their communities with the knowledge gained from higher learning.

“Graduating from college is a huge accomplishment…only about 15 percent of the U.S. population has a bachelor’s degree,” Huston said. “I couldn’t have asked for better preparation for where I’m at now.”

Computer technology gave Huston a challenge, providing an industry in the constant throes of growth and change.

Much of his time at Texas State was spent in the Mitte Honors Program, where he chose to write a thesis on biometric security models. Biometric identifiers include fingerprint, retina scan and other biologic processes that are unique to an individual and difficult to fake.

His study focused on voice identification in hidden markov models (HMM), statistical models that can predict the probability that your voice matches a previously stored electronic version, and testing the reliability of an enrollment phase for speaker verification systems.

Working in tandem with Huston was Wilbon Davis, chair and professor in the Department of Computer Science at Texas State. Huston said that having a relationship with a professor such as Davis makes far more impact than any class ever could.

“Evan is clearly an exceptionally bright young man and will do his alma mater proud,” said Davis. “I'm sure he'll lead a life of accomplishment and influence in many areas in addition to computer science.”

In order to better understand the commercial side of the high tech industry, Huston supplemented his undergraduate education with a trip to Palo Alto, California. He enrolled in the Stanford School of Business summer institute, immersing himself in a curriculum laying out the basic foundations of free enterprise.

He rounded out his undergraduate experience by completing an internship with IBM’s Extreme Blue program in North Carolina. The program allowed him to present his work at conferences in front of hundreds of researchers, scientists and executives.

“When I interned there, I realized that most of their executives in the technical groups had Ph.D.s instead of MBAs,” Huston said. “That’s when I realized a Ph.D. can open any door.”

After graduation, Huston worked for Blue Fish Development Group, a small Austin-based software consultancy firm. He said it made a great impact on his future aspirations.

“I got to work on projects for several Fortune 500 companies,” Huston said. “It really made an impression on me as far as the level of problems I want to work on and the maturity at which I solve them.”

One year later he applied to Rice University, “talking his way” into the doctoral program in computer science.

As a current teaching assistant, the 25-year-old plans to receive a master’s next year on his way towards a Ph.D. His doctoral thesis, while not specifically decided upon, will involve bioinformatics and systems biology. Huston currently teaches courses on the basics of access databases.

“Being a TA puts you in a unique position…because you fall somewhere in the middle of classmate and professor,” Huston said. “Several events this semester have transformed my role as TA to include counselor, friend and mentor.”

Two broad areas that Huston’s research focuses are developing computer models for cellular networks and performing DNA analysis.

In a cell, there is a complex communication network that determines a response to external stimuli.

For example, a faulty cell may receive a signal to self-destruct. Normally, the network for that cell interprets this signal correctly and does so.

However, in a cancerous cell, the signal to self-destruct may be interpreted as a command to reproduce. Cellular networks have millions of nodes, with only a very small portion of their communication behavior understood.

Due to the massive size of cellular networks, there is a need for computation approaches to reconstruct and simulate the exchange of signals involved. The communication network in each person and each cell is different, explaining why cancer patients respond differently to the same treatments.

Huston and his fellow researchers produce statistical measures for communication networks and collaborate with medical scientists to properly interpret results.

The grand goal, one that is likely decades away, is to be able to analyze each patient and develop targeted drugs on an individualized basis based on the unique molecular makeup of each person.

“We consider this a much better alternative to chemotherapy which is a ‘kill everything and see what happens approach,’” Huston said.

His other research focus is using computers for DNA analysis to prevent disease.

More than 99 percent of the DNA in every human on earth is identical, leaving the remaining less than one percent to differ individuals’ genetic traits in areas such as eye color. Many diseases are also genetic, including heart disease, depression and diabetes.

Computational methods for analyzing DNA from various populations can identify the strains responsible for these genetic traits. Upon collaboration with biologists that research ways of applying DNA mutations, it may one day be possible to suppress portions of DNA responsible for genetic diseases and remove those diseases from the human population.

The close proximity to the Texas Medical Center in Houston has resulted in opportunities to work with researchers from Baylor Medical School and the M.D. Anderson Cancer Center in both endeavors.

In addition, the Bioinformatics group at Rice has provided scholarly support for Huston’s progress.

“It is vital for researchers in medicine and computer science to work closely on these projects,” Huston said. “The biologists we've worked with are usually amazed with what we can achieve computationally on these problems.”

When he’s not breaking down the barriers of the human genome, Huston strums his favorite musical instrument, the guitar. He has picked for 13 years, spending at least an hour playing and writing music every day. From 2000-2005, he performed more than 400 sets, even playing with groups such as tthe Black Crowes, Blues Traveler, Bush and Blue October.

“It balances and centers me,” Huston said. “It’s important to have a multi-track life.”

Although often too busy to get away from his studies, Huston says that he misses the serenity and picturesque atmosphere of San Marcos and looks forward to visiting.

“It’s so easy to escape the town and get to the outdoors (in San Marcos),”a reminiscing Huston said. “I miss the people, the square, the river and the simpler life a smaller city provides, including less traffic.”

Huston plans to complete his master’s in computer science next year and begin work on his doctoral thesis shortly after.