Dr. Will Gilbert
Feats of Bioinfomagic
Genes on HTML
When he was satisfied with the trial, Gilbert wrote a PERL script to automate the 192 tests. “Using the vectorized version of BLAST, the Mac had no problem cranking through the 192 comparisons. The whole test took about three hours and we rigged it to produce HTML files for the lab scientists. A/G BLAST saved our cookies.”
At the end of the test, Gilbert had identified five categories of possible genes that make humans human. “A large portion of those genes,” he says, “are involved with neural/nerve development. That’s not surprising; our brain development is one characteristic that sets us apart from other species.”
Because the test results have been streamlined into 192 HTML files, scientists can quickly examine test results with their browsers — and even click on a link that takes them to the annotated genes on the NCBI website. “They don’t have to ferret through a 200-page printout, type in a word and then go looking for it,” Gilbert says. “That was the beauty of the PERL script. You can just hand the 192 HTML files to the scientists and they can go home with their iBooks and review them.”
“Diseases will be detected and treated long before symptoms occur, and therapies designed according to a person’s genetic makeup.”
Maps to Medicine
Gilbert’s work helps to blow open the field of comparative gene research. For researchers who are sequencing the chimpanzee genome, it provides a useful starting point. “Once we get the sequencing for the chimpanzee, we can begin comparing the sequences, and see where a gene is decrepit in one genome and finely tuned in the other. We’re no longer saying ‘I want to compare two organisms that are vastly different.’ We’re starting to look at ‘Here’s a chimpanzee. Here’s a human. How close are they? What are their differences?’
“Or we can start to compare one bacteria with a relative to that bacteria. ‘What makes one cause disease and the other not cause disease?’ These discoveries promise a new era in medicine, where diseases are detected and treated long before symptoms occur, and where therapies are designed according to a person’s genetic makeup.”
Empowering Scientists
Gilbert’s project also empowers scientists. “Right now we’re in the very early years in bioinformatics on Mac OS X,” he says. “I have run A/G BLAST on one of our Mac OS X Servers here at the Genome Center. But it could just as easily have run on a professor’s iBook. And there will come a time very shortly when you will run BLAST just like you’d run Excel.” Pushing himself back about three feet from his desk, Gilbert looks at his Mac and Apple Cinema Display. “I used to have a pile of equipment off to the side that ran other stuff. But, with Mac OS X, I can have email, word processing — this morning I was doing a little video editing — and I have a very robust version of UNIX that can run most of the scientific applications.
“And A/G BLAST is the tool of the post-genomic era. It will analyze larger quantities of DNA sequences at a time — and take us closer to understanding the hereditary instructions that make each of us unique.”
