Lincoln vs. Worm
Image ©2002 Julian Cash

Stein on Bioinformatics


Lincoln Stein was a hospital pathologist until he saw the light and joined the bioinformatics revolution. He learned the tricks of the trade in Eric Lander's lab at the MIT Genome Center before striking out on his own at Cold Spring Harbor Laboratory. He is now an associate professor at CSHL and runs a lab, the main interest of which is in integrative databases for biological sciences.

Lincoln will be a keynote speaker at O'Reilly's upcoming Bioinformatics Technology Conference, February 3-6, in San Diego, California. In his keynote address, he will discuss the challenges of bringing bioinformatics into the biological mainstream, and he charts a course for those who wish to make bioinformatics their future careers.

In this interview, Bruce Stewart talks with Lincoln about the technologies and tools in bioinformatics that he's most excited about.

Bruce Stewart: What cool new genomics technology do you think will most affect, or open, opportunities in bioinformatics?

Lincoln Stein: Nanotechnology. It enables us to formulate and test hypotheses directly by constructing tiny machines to reproduce what enzymes, nucleic acids, and other macromolecules do within the cell. By interfering in very subtle ways with the operation of the cellular machinery, we will be able to create unparalleled models of metabolic processes.

Stewart: What's your opinion on BioPerl? BioJava?

Stein: Wonderful projects, but let's not forget BioPython and BioRuby. All four of these projects have done an amazing job in creating high-quality open source software and have together saved the world thousands of man hours of effort.

Stewart: We've heard a lot about new technologies for ultra-rapid, whole-genome sequencing. Once these technologies are in place, how will that impact our ability to do comparative genomics?

Stein: Right now we think of comparative genomics as comparing two organisms. Once we have dozens or hundreds of organisms to compare we'll need to develop a whole new family of algorithms in order to integrate the data and find significant patterns of difference. It will be extraordinarily exciting.

Stewart: What new tools will be needed to analyze the data? How will requirements change for storage, processing power, hardware, and so forth?

Stein: We'll need bigger disk drives and more processing power, of course, but the biggest change will be the paradigm shift as we need to integrate data through deep sets of evolutionary clades. It will also refocus our attention on the differences among individuals in the same species. We are too used to thinking of the human genome. Soon we'll have to think about handling many genomes.

Stewart: Several universities offer graduate programs in bioinformatics. What do you think of them? What do you think of the quality of students/graduates coming in to your lab?

Stein: The students who have come to my lab from bioinformatics programs in the U.K. have been excellent. I haven't seen any from American universities yet--they must still be in the pipeline. I'm very much looking forward to it.

Stewart: What tool or program do you currently wish you had in your lab today?

Stein: I'd like a reliable, threaded version of Perl.

Stewart: What hardware do you use?

Stein: Primarily Linux running on Intel hardware, with a few Sun Microsystems and Compaq/DEC machines thrown in.

Stewart: What commercial programs do you use in your lab?

Stein: The only commercial programs we use are the Oracle DBMS, Deneba Canvas, and Microsoft PowerPoint. Everything else is open source.

Stewart: How many O'Reilly books do you own?

Stein: Several dozen. I used to own them all, but I can't keep up with them any more.

Return to