This winter, the Virginia Bioinformatics Institute at Tech hired Reinhard Laubenbacher from New Mexico State. He holds a joint appointment as Professor in the Mathematics Department.
Alumni Newsletter: The term “bioinformatics” has made its way into the popular press. Exactly what is bioinformatics?
Reinhard Laubenbacher: There are many different answers. Until recently, biology was data poor. Then technological advances turned biology into a digital science with the challenge of storing, managing, and analyzing very large data sets.
AN: How large?
RL: The human genome has about 3.2 billion base pairs. Think of it as a string with more than three billion characters.
AN: It sounds like you’ve described a job for computer science. How does mathematics fit in?
RL: From my point of view, computer science is used to convert all of this data to information. Mathematics is needed to convert information to knowledge. My role is to design models from which we can extract knowledge.
AN: Can you give us an example?
RL: Biologist were very surprised to discover that the number of genes in the human genome is roughly the same as the number in the mouse genome. We’re pretty confident that people are more sophisticated and complicated than mice, so how can this be? Think of two large networks with the same number of nodes. What makes one network more interesting than the other is found in the connections among the nodes. Molecular biologists have made lots of progress determining the local properties – bits and pieces – of the “genome network”. Mathematicians offer general tools to understand the global picture.
AN: What sorts of general tools?
RL: Everything: graph theory, geometry, topology, differential equations…
AN: That’s exciting. How did you get involved with bioinformatics?
RL: I wrote my Ph.D. thesis on K-theory in 1985. This is a very deep subject, but I was stymied by the difficulty of computing examples. During a visit to Cornell, I learned about the new field of computational algebra and got hooked.
AN: Do you use computational algebra in your bioinformatics research?
RL: The story is much more interesting. In 1995, this guy walks into my office needing help with a computational problem. He is the Director of the Physical Sciences Laboratory at New Mexico State. Well, the problem has nothing to do with computational algebra and I don’t know what to tell him. But his problem intrigued me – suppose you are given an enemy force and you want to anticipate their action. You can monitor their communications traffic but you can’t read any of the messages. However, information about the communications infrastructure the enemy uses is available. Can you get clues about their intentions from the pattern of traffic?
AN: You’ve described a prototype for many different network and communications problems.
RL: Yes. There are lost of variations of this network problem. Many require the development of algebraic measures for patterns of interaction. For example. Drugs are smuggled into the U.S. by planes landing on small airstrips. You know where many of the runways are, but not all. An you use radio traffic patterns to detect the location of hidden airstrips. Here is one of my favorites. Psychologists want to know which factors predict future academic success for young school children. A large amount of observational data is collected that describes how kids play with each other. My problem is to extract patterns from the data to determine whether peer interaction is one of these success factors.
AN: I know you have very strong interests in the history of mathematics. [Dr. Laubenbacher has written a book with David Pengelly entitles Mathematics Expeditions: Chronicles by the Explorers.] Is this perspective useful for research?
RL: History of math is my hobby. I guess it led to a mathematical mid-life crisis. Most Ph.D.s in pure mathematics have a very narrow picture of their profession. A career is a collection of theorems that one has proved. History does give perspective. After years of being uninterested in applications, I wanted to connect to the world. Then one of my former students got involved in a genome project that led quickly to the development of new drugs. What a powerful ideal – a context for mathematics that helps people!
AN: You are obviously excited by this change of direction.
RL: It’s not so big a change. I apply pure mathematics. I can have my cake and eat it, too.
This article was published in the Summer 2002 Mathematics Newsletter.
June 30, 2002